This makes accurate content essential. If our content is stale, even the best agent or search engines will return wrong answers, which is frustrating to everyone.

“Knowledge management today is about making sure people can find the right answers the moment they need them. When knowledge stays current, employees get unblocked faster and productivity improves, making the overall support experience far more efficient.”

Silvina Olkies, senior director, Service Management, Microsoft Digital

That means the knowledge bases that get tapped for answers need be accurate, and for that to happen, they need to be updated frequently and without delay.

Internally here at Microsoft, that’s where our team got involved.

We’re Microsoft Digital, the company’s IT organization, and our team saw an opportunity to use AI to dynamically and proactively update our knowledge management systems.

Our first step was—in partnership with our Global Help Desk—to strengthen our self-serve help and reduce the number of steps users need to take to find the right answers. It’s a process that many of our own customers can apply to their knowledge management transformation.

“Knowledge management today is about making sure people can find the right answers the moment they need them,” says Silvina Olkies, a senior director of Service Management in Microsoft Digital. “When knowledge stays current, employees get unblocked faster and productivity improves, making the overall support experience far more efficient.”

The challenge: Fragmented knowledge, manual reviews

For our Global Help Desk, the challenge wasn’t just the volume of content, but also its condition. Support knowledge is spread across thousands of self-service articles, agent-facing systems, and SharePoint sites, all constantly evolving.

“If the knowledge isn’t accurate and current, the experience breaks down immediately. Bad content leads to bad answers.”

Kevin Verdeck, senior IT service manager, Microsoft Digital

In today’s fast-changing AI-powered world, it doesn’t take long for knowledge to become incomplete, out of date, or redundant. This shows up in the inaccurate answers employees might receive.

In an environment increasingly powered by search and AI, weak knowledge equals weak results.

“If the knowledge isn’t accurate and current, the experience breaks down immediately,” says Kevin Verdeck, a senior IT service manager in Microsoft Digital. “Bad content leads to bad answers.”

At our Global Help Desk, keeping that content current required a manual review process.

Our teams analyzed usage data, depended on support agents to report missing or outdated content, and worked through recurring review cycles that relied on subject-matter experts to help confirm whether articles were still accurate. This took significant time and coordination, and even then, some issues were identified only after employees had already hit a dead end.

The result was a system that was reactive and hard to scale.

When employees couldn’t find answers, issues were pushed to advanced support. Poor knowledge quality created poor outcomes, while those responsible for fixing it were struggling with maintaining it.

“One five-member team was reviewing 1,900 self-service KB articles and 1,700 agent-facing KB articles every six months, and that didn’t even include the many SharePoint sites,” Verdeck says. “It was basically their full-time job doing regular reviews.”

Turning raw data into knowledge

Our team in Microsoft Digital set out to build AI for Knowledge Management, a centralized system that could scale across multiple repositories, cut the manual work of keeping content current, reduce reliance on busy content owners, and prepare knowledge for people and AI to use.

“When you have a large volume of data, it’s a silent gold mine. The sheer brilliance lies in taking that data, making it sing, and letting it tell you exactly where the treasure is.”

Ankit Guddewala, software engineer II, Microsoft Digital

An AI pipeline solution made sense because we wanted to fix the issue at scale.

The real opportunity was to turn every resolved support ticket into a signal that looked at what the employee was asking for (nature of the issue or request), whether the answer was already documented, and how the AI and human agents handled it. So, we began with our large amounts of support ticketing data and systems as the starting point.

“When you have a large volume of data, it’s a silent gold mine,” says Ankit Guddewala, a software engineer in Microsoft Digital. “The sheer brilliance lies in taking that data, making it sing, and letting it tell you exactly where the treasure is.” 

The stages to complete the work happen as follows:

1. Ingest the raw support data: The team pulls in large volumes of incident data from our ticketing systems.
2. Clean and structure noisy data: Tickets can include conversation notes, incomplete details, inconsistent writing styles, and abandoned issues. We use the AI enrichment layer to turn those details into structured fields, such as reported versus actual problems and remediation steps.
3. Find patterns across incidents: We cluster tickets to help identify recurring issues and avoid cluttering the knowledge base with one-off scenarios.
4. Compare patterns against existing knowledge: We use the system to search current articles and rank how closely the resolution aligns to current content to determine what steps to take next. For example:
   • Below 40 percent: Create knowledge
   • 40 to 80 percent: Update existing knowledge with missing details
   • Above 80 percent: No update needed
5. Notify the appropriate knowledge managers: Subject-matter experts are notified by email so they can validate the change(s) and add more detail if needed.

“Our goal is to free people from the manual work of maintaining content so they can focus on improving its quality. With the right human-in-the-loop balance, AI can do the heavy lifting while people make sure the final knowledge is accurate and useful.”

Namrata Ladda, product manager II, Microsoft Digital

The result is far less time spent reviewing thousands of articles during every review cycle. We keep humans in the loop to validate the output. Their feedback helps tune the AI model, so it improves over time.

“Our goal is to free people from the manual work of maintaining content so they can focus on improving its quality,” says Namrata Ladda, a product manager in Microsoft Digital. “With the right human-in-the-loop balance, AI can do the heavy lifting while people make sure the final knowledge is accurate and useful.”

Impacts and what’s next on the journey

Using our new AI for Knowledge Management platform, our Global Help Desk teams can identify knowledge gaps without waiting for someone to report them. They’re using AI to generate structured article drafts so humans can focus on quality, not search through data.

The Global Help Desk projects the solution will save them an estimated 16,000 hours annually; result in a 10 percent reduction in support tickets; and reduce the number of advanced support escalations. This leaves everyone on the team more time to directly help employees more quickly, when needed.

“Turning our knowledge base from a static thing into a living knowledge base is a big step forward,” Ladda says. 

Other Microsoft teams, including HR, have expressed interest in leveraging the content management platform. Once the product has completed internal testing, AI for Knowledge Management will be released to all company employees and customers.

“This solution moves knowledge management from manual maintenance to an intelligent capability that helps organizations scale and apply what they know more effectively,” Olkies says.

The post AI for Knowledge Management: Keeping support content up-to-date at Microsoft appeared first on Inside Track Blog.

At Microsoft, this shift is grounded in years of investment in our neurodiverse employees. From launching our Microsoft Neurodiversity Hiring Program to building a company-wide focus on neuro-inclusive employee experiences, we bring our lived experience to how we design and build our products.

We set out to better understand where our products could do more for all our users and challenge established accessibility norms.

“When we create space for neurodivergent voices to shape our products, we don’t just improve accessibility, we build better technology for everyone. Accessibility is strongest when it is informed by lived experience, and initiatives like our product feedback sessions make sure those voices directly influence how our products evolve.”

Neil Barnett, chief accessibility officer, Microsoft Accessibility

Neurodiversity Celebration Week became the launch point for a new approach to product feedback and co-design at Microsoft. We used the event to create an ongoing program that brings the perspectives of neurodivergent employees directly into the product development process, especially individuals who experience challenges with executive functioning tasks such as focus, learning, memory, organization, and task completion.

For our first feedback round, 150 neurodivergent employees evaluated 10 products across multiple divisions using real-world scenarios. Rather than focusing only on traditional accessibility checks, they shared insights on how products support everyday thinking, learning, and productivity.

The feedback provided product teams with a deeper understanding of where experiences can be simplified or made more intuitive and easier to navigate. More importantly, it established a repeatable process for gathering perspectives from people across the cognitive spectrum, helping teams build products that work better for more people.

“When we create space for neurodivergent voices to shape our products, we don’t just improve accessibility, we build better technology for everyone,” says Neil Barnett, chief accessibility officer in Microsoft Accessibility. “Accessibility is strongest when it is informed by lived experience, and initiatives like our product feedback sessions make sure those voices directly influence how our products evolve.”

How neurodiversity can help a broader audience

Neurodiversity refers to the idea that neurological differences—including autism, ADHD, and dyslexia—are a natural part of human diversity. It recognizes that there’s no single “right” way for the brain to work.

“Neurodivergent employees bring an innovative way of thinking and have ideas that help make our products better.”

Tarena Shanaberger, senior PM, Microsoft Accessibility

Product groups at Microsoft value feedback from our neurodiverse employees. However, without a consistent way to get it, product groups had to reach out with ad hoc requests through the different inclusion networks at the company.

We saw an opportunity to create both lasting change and more efficiency in this process.

“Neurodivergent employees bring an innovative way of thinking and have ideas that help make our products better,” says Tarena Shanaberger, a senior PM on the Microsoft Accessibility team.

Designing with lived experience

Neurodivergent inclusion is most valuable when it’s built in early in the product lifecycle, preferably before assumptions take hold and shape the product in ways that are harder to change later.

“Lived experience is a genuine form of expertise. You can have telemetry, design reviews, usability metrics, and KPIs that mean things to people who review data, but there are aspects of cognitive load and human behavior that become much more visible when you involve people who experience systems differently.”

Karl Niblock, architect, Engineering and Architecture Group Security

Karl Niblock, an architect in our Engineering and Architecture Group Security team, believes that early intent can result in experiences that delight users by making products “shockingly” easy and a joy to use, even for something as routine as logging into a tool.

Designing systems that help reduce unnecessary cognitive friction means more people can do their best work consistently and with more confidence. Those are the types of human-centered insights that aren’t possible to derive from data alone.

“Lived experience is a genuine form of expertise,” Niblock says. “You can have telemetry, design reviews, usability metrics, and KPIs that mean things to people who review data, but there are aspects of cognitive load and human behavior that become much more visible when you involve people who experience systems differently.”

When people share their experiences, everyone can learn and benefit.

“One of the patterns I often see is dead-end workflows, where a user follows the process exactly as instructed but ends up stuck, with no obvious next step,” Niblock says. “For someone with dyslexia or autism, the challenge is often not the task itself but the repeated effort of decoding similar-looking instructions and trying to determine what went wrong. By designing clear recovery paths, plain-language guidance, and visible next actions, we can dramatically reduce cognitive friction. Those improvements help neurodivergent users, but they also make products easier and less stressful for everyone.”

Building a better path for product feedback

Software engineer Jordan Cowe surfaced an idea to host an annual neurodiversity bug bash to uncover where products pose challenges for neurodivergent users. He explains that neurodivergent employees are often some of the first people to identify points of friction before release and says that the bug bash helped shine attention on these issues and encourage product group action on the feedback.

“If something affects a neurodivergent employee, it likely also affects many people who don’t identify as neurodiverse. Fixing these issues improves the products for everyone, making them easier to use and better at keeping users engaged.”

Jordan Cowe, software engineer II, Copilot Engineering team

This bug bash led to a structured way to bring in testers, distribute feedback results to product groups, and infuse key learnings into our products. Our testers asynchronously went through the instructions, followed the scenarios, recorded their screens, talked through their pain points, and completed a product survey. We used the results to rate product usability, which helped us identify issues and ideas for improving our products.

“If something affects a neurodivergent employee, it likely also affects many people who don’t identify as neurodiverse,” Cowe says. “Fixing these issues improves the products for everyone, making them easier to use and better at keeping users engaged.”

What feedback looks like from the product side

Microsoft AI UX researcher Audrey Aday used feedback from the co-design sessions to explore what makes an AI response feel inclusive, useful, and manageable. Tester feedback showed that the issue wasn’t capability, it was control.

“Instead of treating accessibility as a check-the-box exercise, we’re building relationships with neurodivergent employees and embedding their feedback into the full product development lifecycle. We’re moving away from ‘test this for us once’ to ‘build this with us, continuously.’”

Audrey Aday, UX researcher II, Microsoft AI Design

Testers consistently said Copilot can feel overwhelming when it returns too much information at once. They want more control (not less capability) over how much detail they see. In response, product teams are exploring custom instruction menus, smarter defaults, and adaptive personalization that learns individual preferences without creating extra work.

Product teams across Microsoft see lived experiences as a quality driver, not an edge case. Neurodivergent employees help teams spot usability opportunities that benefit everyone, especially around information overload, pacing, and clarity.

“Instead of treating accessibility as a check-the-box exercise, we’re building relationships with neurodivergent employees and embedding their feedback into the full product development lifecycle,” Aday says. “We’re moving away from ‘test this for us once” to ‘build this with us, continuously’.”

How the Inclusive Tech Lab supports co-design

Support also comes from the Inclusive Tech Lab, where Microsoft design teams work directly with users with disabilities to uncover exclusionary designs and identify new product opportunities.

“The best time to engage with us is when employees are starting something new or revising old designs. The Inclusive Tech Lab helps teams partner with disabled advisors early in the design process.”

Sarah Heinzen, senior designer, Microsoft Design and Research

The lab connects teams with people who bring their lived experience across different scenarios, such as limited mobility, photosensitivity, low vision, and neurodivergence. This co-design approach brings in users early and keeps them involved so their lived experience can shape product decisions from the start.

“The best time to engage with us is when employees are starting something new or revising old designs,” says Sarah Heinzen, a senior designer on the Microsoft Design and Research team. “The Inclusive Tech Lab helps teams partner with disabled advisors early in the design process.”

Building feedback into how we work

While the product teams work on incorporating feedback, we’re hard at work solidifying this continuous feedback loop and adding it to everyday product development at Microsoft.

Partnering with the Inclusive Tech Lab, we’re setting the foundation for making that a reality using:

• One consistent path to feedback. Product teams have a unified way to engage with employees from the disability and neurodiversity communities.
• A participant pool of testers. We’re building a diverse network of volunteers to make sure diverse lived experiences consistently inform product decisions.
• Inclusive ways of working. We’re establishing guidelines to make sure testing scenarios match real user needs.

The expectation is start early, stay engaged, and build with the people you’re designing for.

“Even as a software engineer and someone who’s neurodivergent, accessibility used to feel like something you thought about later,” Cowe says. “Now I pause early and ask: Does this make sense for real users? Is it clear? Is it usable? It’s changed how I build. I’m looking for the small friction points, the things people struggle with, and I’m catching them sooner. It’s made me a better engineer.”

The post Boosting accessibility at Microsoft with help from neurodivergent employees appeared first on Inside Track Blog.

Here at Microsoft, we’re no different.

Complaints about the time and effort required to approve expense reports have been consistent from managers across our organization.

“We continuously heard feedback from our leaders that they were spending too much time approving expense reports,” says Michael He, a senior business program manager in the Greater China Region. “They didn’t know where the potential risk actually was, so they had to review everything in detail.”

At Microsoft Digital, the company’s IT organization, we’ve taken this challenge on by introducing an AI-powered Intelligent Risk Engine.

“The Intelligent Risk Engine cuts through complexity, pointing approvers straight to the expenses that warrant attention. This clears up the noise that has been driving review fatigue, especially at quarter’s end.”

Sangay Wangmo, Microsoft Digital experience director, Middle East and Africa

With this new tool, we shifted approvals from uniform, manual scrutiny to automated, risk-based decision making. This enables faster reviews, reduced cognitive load, and improved compliance outcomes.

“The Intelligent Risk Engine cuts through complexity, pointing approvers straight to the expenses that warrant attention,” says Sangay Wangmo, a Microsoft Digital experience director for our Middle East and Africa region. “This clears up the noise that has been driving review fatigue, especially at quarter’s end.”

Looking ahead, we plan to expand the tool to include automated approvals for low-risk cases, creating a more scalable, intelligent, and efficient process. This will ease the pain for our managers and allow them to focus on their strategic work.

Manual approvals in a complex compliance environment

Across our global enterprise, we handle nearly a million expense reports annually. In regions such as Central and Eastern Europe, the Middle East, and Africa (CEMA), our expense approval processes are shaped by diverse regulatory requirements in many different countries.

“As organizations scale, managers naturally have more direct reports, which means more approvals to process. At the same time, accountability for all compliance still sits with the manager, which adds pressure.”

Kethan Parbhoo, general manager, Central and Eastern Europe, Middle East, and Africa

To take one example, the Middle East and Africa—featuring multiple subregions, languages, and local policy nuances—presents a complex challenge for expense management. Applying a consistent risk lens to every case is difficult.

To ensure they stay compliant, leaders often review expense reports in detail, including verifying receipt accuracy, matching invoice data, and checking supporting information (like attendee lists). This level of review requires substantial time and attention, particularly for managers with large teams who receive a high volume of submissions. This becomes even more time-consuming as groups grow.

“As their organizations scale, managers naturally have more direct reports, which means more approvals to process,” says Kethan Parbhoo, a general manager in the Central and Eastern Europe, Middle East, and Africa region. “At the same time, accountability for all compliance still sits with the manager, which adds pressure.”  

The current tool, MS Expense, which was useful in a pre-AI environment, doesn’t provide an optimal user experience. The process was repetitive and depended heavily on manual validation. Existing tools provide limited support for prioritizing risk or simplifying these tasks, resulting in a similar effort being applied to both low- and high-risk expenses. 

As a result, leaders experience increased workload, slower approval timelines, and continued exposure to potential compliance gaps, despite careful review. 

The four top-level internal pain points of the old approval process can be summarized as: 

• Not knowing where the risk is 
• Approvals take too much time, especially at quarter’s end 
• Too much effort is spent on low-risk, routine reviews 
• Issues are found too late, triggering audits and resubmissions after the fact

The Intelligent Risk Engine is helping us address all of these in a unified, cohesive way.

“The system leverages a combination of AI-based risk checks and policy-driven risk checks. This produces a quantifiable baseline score that allows for easier comparison and risk assessment.”

Eric Carnrite, principal product manager, Travel and Expense

AI-assisted risk scoring embedded in MS Approvals 

The Intelligent Risk Engine that our team developed integrates with the existing MS Approvals system, shifting from volume-based checks to risk-based decisioning. We embed this analysis directly into the workflow.

The risk engine evaluates each expense report against multiple criteria, including receipt matching (which leverages AI and optical character recognition), spending patterns, and policy alignment. It assigns a risk score (1-100) and a risk level—1 at the low end and 5 at the high end—and then highlights specific areas that might require attention. 

“The system leverages a combination of AI-based risk checks and policy-driven risk checks,” says Eric Carnrite, a principal product manager for the Travel and Expense team. “This produces a quantifiable baseline score that allows for easier comparison and risk assessment.”

Expense risk score table

The tool also explains why something has been flagged. This allows approvers to quickly understand where to focus their review and to catch issues early, rather than after the fact. 

“AI is effectively doing the initial assessment that a human would otherwise have to do,” Parbhoo says. “It gives you a strong signal, so you can decide quickly where deeper review is needed.” 

With our new model, low-risk expenses can be reviewed and approved with minimal effort, while higher-risk items receive closer examination. It just makes sense to prioritize our work this way.

“Previously, a $10 coffee receipt required the same level of scrutiny as a $300 invoice, which doesn’t make sense at scale,” Wangmo says. “Leaders are forced to treat everything the same, even when the risk level is clearly different.”

Early returns indicate significant improvements. These include:

• Immediate productivity gains, as approvers stop reviewing all expenses manually
• Reduced rework and late‑stage audit findings 
• Stronger governance at scale, without adding headcount or introducing new manual processes

“It pulls the three parts together: Employees, approvers, and auditing and compliance. It will ultimately make it more proactive for all parties involved, rather than reactive—making sure the whole flow of the expense process is more meaningful.”

Michael He, senior business program manager, Greater China Region

Future direction: Expanded automation and standardization 

This kind of AI-powered technology will eventually allow us to pull everything together in one unified system, meeting the needs of all the major players in the expense management process.

“It pulls the three parts together: employees, approvers, and auditing and compliance,” He says. “This will ultimately make it more proactive for all parties involved, rather than reactive—making sure the whole flow of the expense process is more meaningful.”

And we’re not done innovating. The current risk engine implementation establishes a foundation for further automation, and that’s where we’re headed.

One planned enhancement is the automatic approval of low-risk expenses, subject to compliance approval. This could produce significant savings and greater efficiency across our organization.

“With auto-approvals, we’re not talking about a nominal amount,” Carnrite says. “At this point, we’re targeting up to 75 percent of expense reports being automatically reviewed and approved. This could save us around 150,000 to 200,000 person-hours a year—and that’s at the manager and director level.”

We’ve also added advanced optical character recognition (OCR) technology into our expense tools. This now allows for automatic categorization of expenses, so employees don’t have to enter the category manually for each item.

“The end goal is an AI agent that can proactively create an expense report for you and ask you to review it. You would just validate it and move it forward, instead of building it from scratch.”

Salvador Segura, director of business programs, Field Capability Services

Additional future improvements could include expanded use of AI for data validation, receipt processing, and identification of inconsistencies across submissions. Over time, the goal is to support a standardized approval framework that adapts to regional differences while maintaining consistent risk evaluation and reducing manual workload.

At the next level, we’re hoping to use AI to eventually fully automate the creation of expense reports as well. This would be essentially the Holy Grail for this function.

“The end goal is an AI agent that can proactively create an expense report for you and ask you to review it,” says Salvador Segura, a director of business programs in Field Capability Services. “You would just validate it and move it forward, instead of building it from scratch.” 

It’s this kind of AI-powered work environment that we’re pushing for at Microsoft Digital as we play a leading role in our company’s ongoing Frontier Firm journey.

The post Simplifying expense approvals at Microsoft with AI-powered risk assessment appeared first on Inside Track Blog.

To make this ideal a reality, our team in Microsoft Digital—the company’s IT organization—created an agentic AI assistant that we’re now using to connect new devices to our network. We built the agent into GetConnected, the internal portal that 18,000 of our employees, vendors, and network administrators use each month to register their new devices to our network when they turn them on for the first time.

Our new workflow is simple, fast, and intuitive—and it’s a significant step up from our previous experience.

Creating the GetConnected AI assistant is part of the role we play as the company’s Customer Zero, where we test and use our technology and platforms first and then share our lessons learned with customers. In this case, we’re sharing how we used the Microsoft Agent Framework (MAF) to enhance onsite device management for our employees. GetConnected does not apply to remote device registrations.

“We’ve been looking across our set of services and capabilities to find places where we can do some experimentation leveraging AI. We wanted to be able to test our hypothesis around those AI investments and then be able to double down if it proved correct.”

Jason Thompson, principal PM manager, Microsoft Digital

Improving a highly trafficked internal tool using AI

Our employees use GetConnected to ensure their wired and wireless devices are registered on the network, as well as to extend device expiration dates and check on the status of their devices.

Heavy employee traffic and the repetitive actions users tend to take on GetConnected led us to realize that the tool was the perfect candidate for an agentic transformation. Our goal was to turn what was a five- or six-step process into something that could be completed in just one or two actions.

“We’ve been looking across our set of services and capabilities to find places where we can do some experimentation leveraging AI,” says Jason Thompson, a principal PM manager in Microsoft Digital. “We wanted to be able to test our hypothesis around those AI investments and then be able to double down if it proved correct.”

The team decided to start small, focusing on a couple of the most popular and crucial functionalities within GetConnected.

“We’d seen previous projects that were very ambitious fail because they tried to achieve too much at one time. Based on customer feedback, we noticed that registration is the simplest, most common action that users were having trouble with. So we said, ‘Let’s do that first.’”

Aayush Dave, product manager, Microsoft Digital

It was also important to listen to our employees—our Customer Zero frontline users. They told us which actions in the experience mattered most to them.

“We’d seen previous projects that were very ambitious fail because they tried to achieve too much at one time,” says Aayush Dave, a product manager in Microsoft Digital. “Based on customer feedback, we noticed that registration is the simplest, most common action that users were having trouble with. So we said, ‘Let’s do that first.’”

Next up was deciding how the agent would appear in the portal. The Microsoft 365 Copilot model of a sidebar chat menu worked well for other workflows, so it seemed appropriate to approach the new GetConnected experience in a similar way. This enabled us to create a new experience alongside the existing workflow, so customers could still access the original process (should they need to) and compare the two experiences.

“Other teams might decide to automatically replace the UI with an agent,” says Faris Mango, a principal software engineering manager in Microsoft Digital. “But that’s hard, because now you’re forcing people to use the agent. If it’s not ready to be used at full capacity, they don’t have an alternative to accomplish what they intended. We wanted to avoid that situation.”

Testing out Microsoft Agent Framework (MAF)

To build the agent, we considered two paths.

The first was to directly call the Model Context Protocol using JavaScript, an option that would require significant amounts of coding on our part.

“Some declarative agent systems do all of the things in the background, and you don’t get to turn all the little knobs. MAF gives you the flexibility to make the experience exactly what you want.”

Darron Sullivan, principal software engineer, Microsoft Digital

The second was to use Microsoft Agent Framework (MAF), which proved to be simpler and more customizable for our needs.

“Some declarative agent systems do all of the things in the background, and you don’t get to turn all the little knobs,” says Darron Sullivan, a principal software engineer in Microsoft Digital. “MAF gives you the flexibility to make the experience exactly what you want.”

The tricky part, however, was that MAF was fairly new at the time. In fact, the week that the team started developing the GetConnected agent was the same week that MAF was released in preview internally. As we were building out our agent, the framework was going through its own updates, which threatened to hinder our progress. Even one small change to the framework could break our tool’s entire functionality.

“They were moving really fast, and we were adopting new features and finding new bugs all the time,” Sullivan says. “You had to go through that rapid iteration and development, which is a challenge, but it was also pretty awesome because we’re working on the cutting edge.”

The upside was that we were able to provide valuable feedback to the MAF engineers, which in turn could supercharge the work we were doing on our agent. As a bonus, our partnership drove other teams to pursue similar projects.

“The knowledge sharing across our org was notable and crucial,” Dave says. “Our team was one of the first to start building a solution like this, and we presented in numerous architecture forums to share the components and frameworks we were using, and the teams we were working with. This brought the tide up for all boats in our organization, encouraging other teams to start kicking off similar projects as well.”

Building a seamless, discoverable interface

The agent currently has several key functions, the most prominent of which is to register a device on your behalf.

Previously, employees would have to fill in a long, complicated form that asked for a lot of technical details that they often didn’t know offhand, like type of device or the preferred network.

Instead of just selecting options and approving, the flow is more conversational. The user can start with a suggested prompt like “Help me register a device,” and the agent will ask for the required information (with examples for each field). If the user isn’t sure about something (for example, how to find a MAC address), they can ask follow-up questions, and the agent will pull in FAQ and help content to guide them.

Once all the required details are collected, the agent can complete the registration on the user’s behalf after the user approves it.

Once it has your approval, the chatbot submits the request and replies whether or not it was successful. Users can also ask the agent to show devices that are expired or will soon expire, then prompt the agent to renew those devices if desired.

Seamlessly integrating the agent into GetConnected required upgrading the existing user interface using Fluent.

These updates were needed to support the AI interface integration. Specifically, we introduced a custom header action to launch the AI side panel. Prior to upgrading, doing this would have required using Coherence components outside of their intended patterns.

“We used Fluent AI components to build the AI interface. This helped ensure a consistent Microsoft look and feel across the experience, built-in accessibility for scenarios like screen readers and mobile usage, and components that are designed for conversational and agent-driven interactions.”

Nathan Chambers, software engineer, Microsoft Digital

To stay consistent with the existing app architecture, we upgraded core dependencies like Fluent UI and Coherence to their latest versions. As part of upgrading Coherence across several major versions, it also required us to move the feedback experience to Centro to align with the updated patterns. We then needed to update other parts of the experience like navigation, FAQ, and release notes to match those newer component patterns.

“We used Fluent AI components to build the AI interface,” says Nathan Chambers, a software engineer in Microsoft Digital. “This helped ensure a consistent Microsoft look and feel across the experience, built-in accessibility for scenarios like screen readers and mobile usage, and components that are designed for conversational and agent-driven interactions.”

While making these upgrades, we ran tests to ensure the experience was accessible—for example, for screen reader users or others who might access GetConnected on their phones.

“You can have an amazing, strong piece of software that is well built and focuses on security. But if you don’t have the traffic or people are not using it, it’s worthless.”

Faris Mango, principal software engineering manager, Microsoft Digital

Next, we wanted the agent to be as discoverable as possible. Without people actually navigating to it, there would be no way to show proof of concept. So, we built it so the agent automatically popped open via a side panel when someone loaded GetConnected.

“You can have an amazing, strong piece of software that is well built and focuses on security,” Mango says. “But if you don’t have the traffic or people are not using it, it’s worthless.”

We also wanted to gather early feedback from users. Before releasing it to the entire company, we had internal team members and frequent GetConnected users give the agent a try. Almost immediately, it was clear we had too many approval notices.

“At the beginning, we would have approvals for every single action. For example, if you wanted to see a device in different regions like Puget Sound, Latin America, or Canada, you had to do a separate approval for each region,” Dave says. “This was a very painful experience. So we removed all the approvals and pared it down to a one-click experience.”

Users also had issues with the approval language, which they said was hard to understand and looked like an error message. The next iteration took out much of the technical jargon, making the message more conversational and easier to read.

An agent experience driven by feedback

Our work as Customer Zero is never done. For GetConnected, we’re eager to keep collecting feedback. One major goal is to improve the agent’s performance, making it faster and more responsive.

Our feedback survey is tied directly to a performance dashboard, which tracks metrics like new and returning users, total unique users, conversions, and interactions. Each user submission generates a work item.

“When users leave feedback about something they don’t like, I feed that to the team, and then we sit down and figure out how we can improve that specific part of the experience,” Dave says.

With each update, we’re seeing the payoff of more users and more interactions. The traditional method of registering a device is also seeing a drop-off as more people lean on the agent for assistance.

“Before, you used to have to go into the system and change something about the experience manually. Now, our engineers are going to an AI model and telling it, ‘Hey, you’re doing this part wrong, please improve it.’”

Aayush Dave, product manager, Microsoft Digital

Looking ahead to more use cases

Building an agent has allowed the team to embrace an entirely new type of engineering.

“Before, you used to have to go into the system and change something about the experience manually,” Dave says. “Now, our engineers are going to an AI model and telling it, ‘Hey, you’re doing this part wrong, please improve it.’”

It’s also serving as a reminder to seek progress over perfection.

“AI is changing things so quickly,” Thompson says. “It’s better to do rapid prototyping and roll it out, and start getting the data in terms of how successful the experience is. Then you can let that guide you, in terms of how you iterate going forward.”

Because of the success we’ve had with the GetConnected device registration feature, we’re already exploring other capabilities, including bulk operations to accommodate our facilities managers who need to onboard many devices at once. As AI agents become mainstreamed in many workflows across our organization, we anticipate usage and functionality will continue to grow exponentially.

The post Simplifying device registration at Microsoft with an agentic AI assistant appeared first on Inside Track Blog.

At Microsoft Digital, the company’s IT organization, we’re using a set of employee councils and connected capability groups to guide and accelerate how we deploy and adopt AI across our enterprise. Our goal is to focus our energy on the AI-enabled scenarios that matter most, reducing duplication, strengthening accountability, and making sure our investments create measurable value.

“Our AI decisions and direction must be grounded in business strategy. AI councils provide guidance and enablement for our organization, ensuring that our investments in AI generate tangible benefits to our business. It’s not just developing technology and then looking for a problem to solve with it—we start with the opportunity.”

Don Campbell, principal group technical program manager, Microsoft Digital

That focus matters, because AI success doesn’t come from usage alone. It comes from connecting strategy, enablement, data readiness, responsible AI, continuous improvement, change management, and measurement into one driving force.

That’s how we’re moving from experimentation to repeatable outcomes and from AI enthusiasm to AI accountability.

“Our AI decisions and direction must be grounded in business strategy,” says Don Campbell, principal group technical program manager in Microsoft Digital. “AI councils provide guidance and enablement for our organization, ensuring our investments in AI generate tangible benefits to our business. It’s not just developing technology and then looking for a problem to solve with it—we start with the opportunity.”

Our council-based approach is helping us accelerate our Frontier Firm transformation. The councils work together to set direction for AI adoption at Microsoft Digital, ensuring that our business needs drive solution development that can keep up with the pace of AI change. This work includes building visibility into all our AI solutions, including agents and Model Context Protocol (MCP) servers, while establishing governance and proven practices; developing training and learning pathways; and connecting teams together that are working on similar solutions across the enterprise.

We’re excited for a future where our employees use intelligent agents and human judgment together to work smarter, move faster, and unlock new value for Microsoft and our customers.

Why we use councils to guide internal AI efforts

Effective AI needs both enterprise guidance and business-owned direction. That’s why we’re using councils and connected capability groups as the operating model for our AI deployment.

Each group has a distinct role, and none of them work alone. Together, they help us connect the strategy for AI to the work currently happening across Microsoft Digital.

• Our strategy council sets priorities by aligning AI work to business goals, identifying top scenarios, prioritizing investments, and keeping KPIs and value in focus.
• Our enablement council uses our AI Center of Excellence to turn strategy into action through technical guidance, proven practices, ideation, learning, knowledge sharing, culture, and governance.
• Our data council strengthens the AI foundation via data strategy, governance, access, quality, literacy, and prioritization.
• Our process council drives continuous improvement through operational excellence, problem solving, prioritization, value realization, coaching, and learning.
• Our compliance council applies Responsible AI principles to ensure compliance, inclusiveness, fairness, transparency, and reliability.
• Measurement ties it all together by tracking both business outcomes and engineering artifacts, ensuring we can clearly demonstrate real-time value realization.  

These councils help us see across the business landscape through the lens of AI. They enable us to reduce duplication, scale what works, and make better decisions about where AI can create value. That allows our teams to keep moving fast without letting activity get ahead of accountability.

Aligning AI strategy to business value

Our strategy council helps us decide which AI-enabled scenarios deserve the most attention, which investments align to our business priorities, and how we’ll know whether the work is creating value. It gives leaders a practical way to look across the portfolio and keep our AI work tied to the outcomes we’re accountable for.

“Business strategy defines the what and the why. AI defines the how, enabling execution of the strategy and delivering real value. We should use AI to advance our business strategy, not the other way around.”

Qingsu Wu, principal group product manager, Microsoft Digital

This is important, because broad experimentation is useful early on in your AI journey. It helps teams learn and build momentum. But experimentation has to mature into focus. Without that shift, organizations can end up with too many different agents, agent skills, MCP servers, and other artifacts, without a clear view of what’s actually impacting the business.

We’re using the strategy council to keep that from happening.

“Business strategy needs to lead the AI strategy,” says Qingsu Wu, a principal group product manager in Microsoft Digital and an influential member of the strategy council. “Business strategy defines the what and the why. AI defines the how, enabling execution of the strategy and delivering real value. We need to use AI to advance our business strategy, not the other way around.”

That principle shapes how we work. We use the strategy council to identify our top AI-enabled scenarios, clarify the value we expect to create with each one, and connect that work to a monthly operating rhythm. Product owners still manage delivery and the council keeps the portfolio focused, visible, and aligned.

Tuning strategy into repeatable execution

Our AI Center of Excellence (CoE) is at the heart of our approach to enablement. It helps us translate enterprise AI priorities into practical guidance and execution support for teams building AI-enabled solutions.

“We can see patterns that a single team can’t. We’re translating AI CoE strategy and enterprise priorities into clear execution plans that work in each organization’s context. That allows us to align priorities and make sure our biggest bets are actually landing.”

Ria Khetan, senior program manager, Microsoft Digital

The AI CoE extends the reach of the strategy council. It gives teams what they need to build, govern, reuse, and scale what matters, while the strategy council assists us in deciding where to focus.

That connective role is central to the broader council model. The strategy council identifies the top AI-enabled scenarios. The AI Center of Excellence connects strategy to execution across the organization, operating as a cross-functional coordination layer that sets direction and creates shared accountability.

“We can see patterns that a single team can’t,” says Ria Khetan, a senior program manager in Microsoft Digital, who is a member of the council. “We’re translating AI CoE strategy and enterprise priorities into clear execution plans that work in each organization’s context. That allows us to align priorities and make sure our biggest bets are actually landing.”

The COE helps teams move those scenarios forward with answers to important questions:

• What initiatives are in flight?
• What initiatives bring the most return on investment?
• Where is there potential duplication?
• Where do we need clearer guidance?
• Where do we need stronger governance?

It also helps reduce fragmentation. When teams build in isolation, they can solve the same problem in different ways. They can choose different patterns, interpret standards differently, or create solutions that don’t scale beyond a single context. Enablement gives us a shared way to look across that activity and ask better questions.

“We use the CoE to bring consistency to how AI work gets done,” Campbell says. “It gives us a way to step back and ask whether we’re solving the right problems and whether we’re set up to scale.”

“High-quality, well-governed data is essential to accelerate AI implementation and adoption, and to ultimately unlock its full value. Data quality, accessibility, and governance are imperatives for AI systems to be reliable, scalable, and business-critical. Recognizing this principle is propelling our data strategy.”

Miguel Uribe, principal PM manager, Microsoft Digital

Building AI on trusted data

Our AI scale depends on trusted and reliable data. That makes our data council central to our council-based approach. This council makes sure our teams work with data that’s governed, discoverable, accessible, and ready for AI.

“High-quality, well-governed data is essential to accelerate AI implementation and adoption, and to ultimately unlock its full value,” says Miguel Uribe, a principal PM manager in Microsoft Digital and member of the data council. “Data quality, accessibility, and governance are imperatives for AI systems to be reliable, scalable, and business-critical. Recognizing this principle is propelling our data strategy.”

We’re applying a data mesh mindset to balance domain ownership with enterprise consistency. Teams stay close to the data that they know best. Shared standards for governance, quality, metadata, and compliance provide a framework to make that data useful across Microsoft Digital.

Microsoft Fabric and Microsoft Purview are key to that approach. Microsoft Fabric unifies our siloed data in a shared data mesh. Microsoft Purview enables governance and best practices to ensure that we manage our data responsibly through discovery, classification, protection, and monitoring.

Our goal is AI-ready data that’s available, complete, accurate, and high quality. Our data council also works with the AI Center of Excellence to strengthen data and AI fluency through learning pathways, operational practices, and community programs.

“Our capacity to drive process improvements has been crucial to our AI transformation as a company. We’ve adopted a ‘CI before AI’ approach to ensure that we don’t end up automating inefficient processes.”

David Laves, director of business programs, Microsoft Digital

Improving the process before applying AI

AI works best when it’s applied to the right problem. That’s why continuous improvement is part of our council-based approach. Before teams automate a workflow or build an agent, we want them to understand the process, identify waste, and decide where AI can create measurable value.

“Our capacity to drive process improvements has been crucial to our AI transformation as a company,” says David Laves, director of business programs in Microsoft Digital and a member of the Continuous Improvement Center of Excellence. “We’ve adopted a ‘CI before AI’ approach to ensure that we don’t end up automating inefficient processes.”

Continuous improvement helps teams make sure the underlying work is worth scaling. That’s when a continuous improvement approach can help. It encourages practices like Gemba walks, Kaizen events, bowler cards, and monthly business reviews that allow our teams to understand where work gets stuck and where AI can help.

Continuous improvement keeps the council model grounded in real work. We’re applying it where the process is understood, the value is clear, and the outcome can be measured.

Scaling AI responsibly

Our compliance council encourages the application of Responsible AI, so our teams can move faster with confidence. As our AI work scales across Microsoft Digital, responsible AI has to connect directly to the same council ecosystem that guides strategy, enablement, data, process, and measurement. That connection helps teams understand what they’re accountable for before they build too far, too fast.

Our responsible AI work focuses on compliance, inclusiveness, fairness, transparency, reliability, privacy, security, and accountability. It’s grounded in the Microsoft Responsible AI Standard and supported by responsible AI champions who help teams apply those expectations in real development workflows.

This approach gives teams structure. It allows them to assess impact, identify risks, document decisions, and bring in the right reviewers. It also creates consistency, as more AI agents and solutions move from experimentation into enterprise use.

The goal is to enable AI project teams to move in the right direction with the right safeguards. Responsible AI gives the strategy council, the AI Center of Excellence, the data council, and product teams a shared standard for trust—to turn ambition into accountable execution. It also makes sure the AI systems we scale are worthy of the trust that employees, customers, and the company place in them.

Measuring our AI outcomes

Our councils choose the right AI work, support teams as they build, strengthen the data foundation, apply responsible AI, and improve processes before we scale. But we still need to answer the most important question: What changed because of the AI investment?

That’s why we have built a common value measurement framework across Microsoft Digital. Our teams use the framework to define expected value before they build. With it, they can establish a baseline, track results, and review what they learn with the right business and AI owners.

We organize AI value across six areas: Revenue impact, productivity and efficiency, security and risk management, employee and customer experience, quality improvement, and cost savings. Not every initiative needs to deliver value in every category. The point is to create a shared language that leaders and teams can use to compare investments, make tradeoffs, and understand progress.

Measurement also pushes us past simple savings claims.

If AI saves time, reduces cost, improves quality, or increases coverage, we want to know what happens next. Did teams reinvest that capacity? Did service improve? Did risk go down? Did quality increase?

AI accountability depends on that full loop. We define value, measure results, review progress, and adjust. Then we use what we learn to guide the next round of decisions.

Operating as one connected AI system

Our AI councils make a difference because each group has a different focus.

“What got us here won’t get us to where we need to go next. We started with broad experimentation—getting teams excited and building—but now we’re evolving as an organization to think about scale, alignment to business goals, and making sure our investments are driving the right outcomes.”

Myron Wan, principal group product manager, Microsoft Digital

Strategy assists us in choosing the right priorities. Enablement helps our teams to build with shared patterns. Data readiness gives AI systems a trusted foundation. Responsible AI allows us to move faster with confidence. Continuous improvement makes sure we’re improving the work before we automate it. Measurement tells us whether the investment changed anything meaningful.

Together, this system means we can operate AI as a business-driven enablement system.

“What got us here won’t get us to where we need to go next,” says Myron Wan, a principal group product manager in Microsoft Digital. “We started with broad experimentation—getting teams excited and building—but now we’re evolving as an organization to think about scale, alignment to business goals, and making sure our investments are driving the right outcomes.”

There’s more work ahead. We need to keep scaling enablement, improving data readiness, increasing high-value use cases, showcasing measurable impact, and tightening alignment across teams.

We also need to keep asking the hard questions: Where should we invest? Where are we reducing risk? Are we reinvesting the value that AI creates?

Our council-based model allows us to answer those questions with discipline. It helps us connect AI ambition to business outcomes and move from experimentation to repeatable enterprise value. And it provides a practical model that other IT organizations can adapt as they guide their own AI deployment.

The post Guiding our AI deployment with a set of employee councils appeared first on Inside Track Blog.

At first glance, helping sellers improve the way they work looks like a data challenge because we tend to believe more data equals better decisions. However, when we explored how they use data to prepare for customer meetings and internal business reviews, we saw a different pattern.

The amount of data they had available to them had increased—but so had the manual effort needed to turn it into useful and actionable insights. In fact, our sellers were spending more time looking for relevant information than they were spending engaging with customers to improve their experience.

As part of our Customer Zero approach at Microsoft, we apply our technologies internally and use that experience to understand what works at scale so we can pass that knowledge onto our customers. In this case, our discoveries unearthed a deeper issue with how our sellers interact with data in their day-to-day work.

Too much data, not enough insight

The increased amount of data available to our sellers gave them more options but also more information to process. It also introduced fragmentation.

“I think the big feedback pain point we got was that we have way more data than our sellers were used to. You give us a gigantic well of data, but how I use that data to execute my job is fragmented.”

Michael Toomey, revenue insights lead, Finance Data and Experience

A significant portion of our sellers’ time went into finding, interpreting, reconciling, and analyzing the information before any actual decisions or customer interactions were taking place. In some cases, sellers were navigating hundreds of reports and dashboards across multiple systems. In others, they were exporting data into spreadsheets, building their own analyses, and assembling presentations manually. They told us they were overwhelmed by the amount of data they had access to.

“I think the big feedback pain point we got was that we have way more data than our sellers were used to,” says Michael Toomey, a revenue insights lead on our Finance Data and Experience team. “You give us a gigantic well of data, but how I use that data to execute my job is fragmented.”

This tedious manual user experience didn’t match the fast pace of our work. Sellers needed to move quickly, and we needed a way to empower them to find the data and insights they needed to be able to make actionable decisions immediately.

Instead of continuing to expand the amount of data available to our sales force, we concentrated on making the existing data easier to access, understand, and use.

Beginning with a trusted data foundation

The starting point was the data itself. Our sales organization draws from many systems, each of which has its own structure, definitions, and refresh cycles. Without alignment across these systems, even seemingly simple seller questions could produce different answers depending on the data source. This created more work as our sellers hunted down which answer was the correct one.

We consolidated all this data into a unified model on Microsoft Fabric, creating a shared data foundation across the organization. It included standardized definitions, consistent metrics, and common hierarchies. Data from more than 70 systems was brought together into one governed environment.

This step required coordination across teams and ongoing attention to evolving data governance. It established trust in our data that made it easier to streamline the user experience for our sales teams down the line.

Streamlining how sellers work with data

With a consistent data foundation put into place, we turned our attention to how sellers could most effectively access and use the information.

Instead of asking sellers to navigate a portal of reports, we designed role-based dashboards that reflect how people actually work. Individual dashboards are curated based on role, workflow, and responsibilities, helping our sellers quickly find what they needed without searching across systems. We transformed the work surrounding actionable insights, too: Power BI surfaces intelligent insights to our sellers’ dashboards every morning, automatically giving them their priorities for the day and providing an overview of their funnel.

Before this transformation, building PowerPoint presentations for customers and summarizing reports were routine but tedious parts of the preparation process for our sellers before they spoke to customers. These preparatory steps have been streamlined or automated wherever possible, reducing seller time spent on repetitive manual tasks.

Empowering sellers using AI

With the data organized and cleaned and the user experience transformed, we introduced AI to overhaul how our sellers interact with the information available to them.

“A lot of what we’re trying to do is get to a native Microsoft 365 Copilot experience where we meet people where they’re working every day. And then work is optimized to be able to reason over our day to pull the data in.”

Michael Toomey, revenue insights lead, Finance Data and Experience 

To find the information they’re looking for, sellers can now ask questions in natural language and receive answers directly in the tools they already use. Insights powered by Power BI Copilot, Fabric, and AI Foundry are delivered proactively, helping sellers prioritize their day, generate summaries, and assemble materials for meetings without working through each step manually.

Insights that previously required a great deal of time to uncover are surfaced directly. The relevant reports are automatically routed to sellers based on their job description and client base. Sellers can also subscribe to continuously updated reports and have them surfaced each morning to their inbox.

These capabilities are available in the tools our people already use, which means they don’t have to spend time learning new product suites. The information they’re looking for can be surfaced through tools like email or within Microsoft Teams, which means sellers can stay focused on their work without needing to switch between systems.

Processes that once took hours can now be completed in minutes.

“A lot of what we’re trying to do is get to a native Microsoft 365 Copilot experience where we meet people where they’re working every day. And then work is optimized to be able to reason over our day to pull the data in,” Toomey says.

What changed

The most visible shift we’ve observed is how sellers spend their time. Less effort goes into finding, assembling, and reconciling information, and more attention is directed toward understanding customer needs and making decisions that actively move the business forward.

After deploying this model across our sales organization, we saw meaningful improvements in both efficiency and effectiveness, including:

• 100,000 seller hours saved annually
• 30% reduction in data ingestion costs
• 10x faster insight generation
• 1,500 reports retired and consolidated

We credit our success to building a trusted data foundation; a streamlined, intuitive user experience; and embedding AI into the flow of work to help our sellers surface insights and information with a few clicks instead of spending hours sifting through irrelevant inputs.

We learned that executive sponsorship and active change management were essential to transforming the department successfully. Concentrating on consistency and departmental alignment around a set of trusted, verified, well-governed data and shepherding people through shifting how they worked with the data were just as important as adding AI into the process.

Looking ahead

At the heart of our trusted data foundation is the semantic layer, the place where our business definitions, metrics, and data quality are standardized across the organization. It’s now the engine that powers our agents. Because those definitions live in one governed place, our agents can reason over our data products with confidence, and we can trust what they surface.

That foundation is already bearing fruit. A new generation of personal and role-based agents at Microsoft is taking on the workflows that were consuming our sellers’ time, such as meeting preparation, pipeline reviews, and customer insight generation. These processes are now running on the governed data infrastructure we built. Everything draws from the same source of truth, so everything our agents work with is consistent, reliable, and ready to act on.

This return on our platform investment is bringing us closer to becoming a Frontier Firm where our people spend less time searching for the answers and more time acting on them.

The post From data sprawl to AI-driven seller insights at Microsoft appeared first on Inside Track Blog.

• Introduction
• Chapter 1: Understanding Work IQ
• Chapter 2: Establishing trust: How we govern Work IQ
• Chapter 3: How our employees experience Work IQ day-to-day
• Chapter 4: Work IQ beyond Microsoft 365
• Chapter 5: Working with Work IQ: The Customer Zero impact
• Conclusion: Where we’re heading: Work IQ, Fabric IQ, and Foundry IQ

Improving agentic workplace results with Work IQ

Adding deeper contextual intelligence to Microsoft 365 Copilot responses

Enterprise knowledge is perhaps a company’s most valuable asset, but for AI and agents, it can be difficult to take advantage of. Years of emails, documents, chats, meeting recordings, and workflows have created enormous volumes of rich data, scattered across systems and teams in a fragmented way. This data captures how work actually happens, but harnessing it broadly—especially in ways that support better decision making—has traditionally been almost impossible.

Enter the power of agentic AI tools.

In the modern agentic workplace, employees and teams here at Microsoft and elsewhere are finally able to take advantage of all that rich, unstructured knowledge. Microsoft 365 Copilot and AI agents can now access all this data and not simply retrieve information but also reason over it—learning how work gets done and then providing rich contextual responses and guidance.

“By giving AI the ability to reason across the vast repositories of unstructured data that our enterprise possesses, Work IQ fundamentally changes what’s possible for Copilot, agents, and employees alike.”

Brian Fielder, vice president, Microsoft Digital

We’ve given this new, dynamic way of leveraging your enterprise data to boost productivity a special name: Work IQ.

Work IQ represents a big step forward.

For us, it’s enabling the concept of “intelligence on tap” across our enterprise, making our organizational knowledge and work context accessible in real time, grounded in the signals employees generate every day. This transforms unstructured data from a challenge into a strategic resource—one that can support workflows at scale.

“Work IQ represents the next phase of the agentic workplace of the future—and it’s here,” says Brian Fielder, vice president of Microsoft Digital. “By giving AI the ability to reason across the vast repositories of unstructured data that our enterprise possesses, Work IQ fundamentally changes what’s possible for Copilot, agents, and employees alike.”

“It’s not really a brand-new capability, but more an evolution of what users already know, which is access to the grounding data in their Microsoft tenant. The difference is that Work IQ adds an additional layer to provide more context, allowing for richer and more relevant results.”

Aisha Hasan, principal product manager, Microsoft Digital

Internally here at Microsoft, Work IQ is having a tangible effect on how we work every day. A few simple scenarios that illustrate the power of Work IQ—described in greater detail in Chapter 3—include:

• Helping our employees understand which emails require their immediate attention, so they can focus on what matters
• Connecting meeting transcripts to the people involved in a meeting, accelerating actions through a deeper understanding of the participants and their work patterns
• Enabling our employees to create, organize, and publish Microsoft 365 content more quickly and with higher quality

This is just the beginning. As AI continues to permeate our business workflows, nearly every day-to-day task at Microsoft will be simplified, expedited, and improved by the intelligence of Work IQ. This includes the agents that are managing routine business and operational processes, giving them critical business context that helps their reasoning abilities.

This guide explores the ways that Work IQ is impacting how work gets done at Microsoft, and how Microsoft Digital—the company’s IT organization—has played a key role as Customer Zero, validating how Work IQ behaves under real enterprise conditions. It also examines the challenges and considerations that IT organizations will face as we enter an era where AI agents have access to unstructured data to complete workflows.

Chapter 1: Understanding Work IQ

Providing deeper insights through the power of context

Before we can fully explore the implications of Work IQ, it’s important to start with a clear understanding of what it is.

Work IQ is not a new application or service that users interact with directly. Rather, it’s a shared intelligence layer that continuously interprets work happening across the tenant. Understanding this distinction is critical, because it explains why Work IQ shows up everywhere Microsoft 365 Copilot works—and why it must be treated as foundational infrastructure, not as optional, add‑on functionality.

“It’s not really a brand-new capability, but more an evolution of what users already know, which is access to the grounding data in their Microsoft tenant,” says Aisha Hasan, a principal product manager in Microsoft Digital. “The difference is that Work IQ adds an additional layer to provide more context, allowing for richer and more relevant results.”

Work IQ is built on three layers:

• Data: It unifies signals from files, emails, meetings, chats, and business systems.
• Memory: It builds persistent understanding of how people and teams work.
• Inference: It combines models, skills, and tools to reason and act.

At a high level, Work IQ consists of the systems that collect and interpret signals from everyday work. These signals come from many familiar Microsoft 365 applications—Word, Outlook, PowerPoint, Teams, SharePoint, and more—as well as structured data sources (such as those contained in Power Apps and Dynamics 365 resources).

The fact that Work IQ unifies unstructured and structured data into a shared ontology is a key differentiator from traditional search tools. This combination, referred to as semantic unification, means that it can combine the authoritative data contained in structured sources with the intent, nuance, and narrative found in unstructured data.

Work IQ draws from a broad range of work data from your Microsoft tenant. The unstructured data includes:

• SharePoint sites, files, and other content
• OneDrive activity that reflects individual work and collaboration patterns
• Teams content, including chats, channels, and meeting data
• Outlook emails and attachments

In addition, calendar signals—such as meeting participation, recency, and frequency—add time-based context that helps Work IQ understand priority and relevance of different data. This is what it means to go beyond simple information retrieval.

When all these are combined, it provides rich context that allows Work IQ to reason across all our employees’ work in a way that would be impossible if each signal were evaluated independently.

In practice, this means that when an employee asks a question about a current work project in Copilot, the tool’s response is not simply informed by the model’s capabilities or general source material. Responses are shaped by Work IQ’s understanding of the employee’s role, recent work, collaboration patterns (who they work with), and the larger enterprise context and conversations surrounding the question.

How our employees interact with and understand Work IQ depends on their role in the organization.

Our personas and their relationship with Work IQ

End user

Gains richer, context-aware agent conversations

IT admin

Enables and governs Work IQ access

Agent builder

Manages data connectors quickly and easily

Security professional

Knows that sensitive data stays protected with Work IQ

AI agents using Work IQ behave similarly. They use the intelligence to ground their reasoning in real organizational data, ensuring that their actions and recommendations are aligned with how work is happening inside the tenant. Although there are differences in how they are configured, all agents in a Microsoft tenant can be set up to take advantage of the power of Work IQ.

The impact of Work IQ on our company has been dramatic—we’re seeing agentic responses and actions that go deeper than surface-level answers. Our ability to reason over both our structured and unstructured data is producing richer, more nuanced contextual results that are boosting our productivity.

As your organization assesses your level of AI readiness, think of Work IQ not as an abstract concept but as critical infrastructure. It’s the key to connecting enterprise knowledge, trust, and productivity in a single, shared foundation.

Work IQ versus Microsoft Graph

Work IQ does not replace what we call the Microsoft Graph, the general term for unified, API-enabling, secure, permission-aware access to Microsoft 365 data, insights, and services. While the Microsoft Graph provides our employees with access to all their work data, Work IQ turns those signals into meaningful context that AI can reason over. In other words, Graph answers the general question “what info exists,” while Work IQ interprets what that information means and weaves it into responses to make them better.

Key takeaways

As you prepare for Work IQ, these points can help frame how to think about its role in your organization:

• Work IQ is foundational infrastructure, not a user-facing feature. It operates as a shared intelligence layer across the tenant, continuously interpreting signals from everyday work.
• Work IQ draws its power from context, not isolated data. By combining signals from email, meetings, documents, calendars, and collaboration patterns, it enables Copilot and agents to reason about work in a way that goes beyond simple search or retrieval.
• Better agentic outcomes depend on Work IQ being in place. When agents and Copilot are grounded in Work IQ, their responses and actions align more closely with real enterprise work, delivering better relevance and measurable productivity gains.

Learn more

How we did it at Microsoft

• Discover how Work IQ is supercharging our AI usage at Microsoft. This post delves into the ways that Work IQ is boosting agentic response and changing the way we work at the company.  
• See how we’re embracing this agentic moment at Microsoft. This story covers how we use the power of Copilot and AI agents to boost employee productivity.
• Explore our IT playbook for the AI era and learn how we’re becoming a Frontier Firm. This article shares our journey as an IT organization toward the AI-powered future.

Further guidance for you

• Learn more about Work IQ with this post from the Microsoft 365 Copilot Blog.
• Check out this Microsoft Learn overview of the different Work IQ agent tools.
• Read all about the Microsoft Graph on the Microsoft Learn site.

Chapter 2: Establishing trust: How we govern Work IQ

Building on an existing foundation of solid governance and security

Like all Microsoft products, Work IQ was built with security foremost in mind. As the IT team at Microsoft, it is our responsibility to work in tandem with the product groups to ensure that all data that Work IQ has access to is well governed and secure.

The fact that Work IQ does not introduce new data into Microsoft 365 makes meeting this commitment easier. Embedded directly into the Microsoft 365 intelligence stack, Work IQ inherits the same compliance, security, and access controls that already govern the tenant.

“With great power comes great responsibility, and it’s up to your IT team to think about what it means to give your users full access to all this Work IQ data. It can greatly accelerate what people can build and what they can do.”

David Johnson, principal PM architect, Microsoft Digital

For Microsoft 365 Copilot–native agents, Work IQ is enabled to provide governed, context‑aware access to Microsoft 365 work data without requiring developers to build or manage individual data connectors.

As our governance experts note, this represents an inherent trade-off. Giving an agent access only to certain isolated data types reduces risk but also limits its value. Granting access through Work IQ means an agent can reason across everything the employee can access. This simplifies enablement but also requires stronger confidence in governance foundations.

Microsoft 365 intelligence stack

As our governance experts note, this represents an inherent trade-off. Giving an agent only access to certain isolated data types limits risk, but it also limits its value. Granting access through Work IQ means an agent can reason across everything the employee can access. This simplifies enablement but also requires stronger confidence in governance foundations.

“With great power comes great responsibility, and it’s up to your IT team to think about what it means to give your users full access to all this Work IQ data,” says David Johnson, a principal PM architect in Microsoft Digital. “It can greatly accelerate what people can build and what they can do. At the same time, organizations will want to think about the downstream implications of access.”

Exposing underlying governance issues

Our overall solution was to anchor Work IQ to our governance and security policies that already existed for our data. Sensitivity labels, data protection rules, and data-loss prevention policies remain the primary guardrails, as they do for all data across our enterprise. All these controls live at the data layer.

A critical aspect of this governance model is how sensitivity labels propagate through Work IQ experiences. In Microsoft 365, the label that is applied to a source document determines the label of any derived outputs, including summaries, insights, or AI-generated responses. This ensures that users have immediate context about the information’s sensitivity and how it should be handled. The label effectively travels with the data, reinforcing both user awareness and policy enforcement.

Labels also play a key role in controlling access beyond simple permissions. Even if a user has baseline access to a location, sensitivity labels can further restrict whether content can be extracted, shared, or surfaced through AI experiences. In some cases, organizations can configure policies so that content with specific labels is not returned at all in Work IQ or Copilot responses. This gives IT teams an additional layer of control to prevent exposure of particularly sensitive information.

These labeling principles extend across collaboration scenarios as well. For example, meeting labels determine the classification of all downstream artifacts—including recordings, transcripts, and notes. Sensitive discussions remain governed consistently, even as Work IQ helps make them more discoverable and actionable.

For example, even with Work IQ enabled, a document labeled Highly Confidential cannot be exposed through Copilot to someone without access, even if it is referenced in a Teams meeting transcript or included in an AI-generated summary. Copilot may understand that the document played a role in a particular decision, but it cannot extract or reveal its contents beyond what permissions allow.

This distinction—discoverable versus extractable—proved critical in our deployment of Work IQ. The intelligence layer makes data relationships visible, but it does not override protection. In one internal scenario, a sensitive document was found to be accessible through a Copilot query. The root cause was not Work IQ, but a missing sensitivity label—the AI tool simply honored what governance allowed. We treated the incident as a governance signal and corrected labeling at the source.

Remember that Work IQ can only access data that:

• Exists inside your Microsoft 365 tenant or is explicitly connected via approved connectors
• The current user already has permission to access
• Is allowed by tenant‑level admin policy, compliance, and sensitivity controls

The security and governance considerations also extend to how new agents are released across our enterprise. For example, an agent created for use within one internal team has lighter governance controls than one that is published to our internal Microsoft agent portal, which offers companywide access. The latter requires additional review, approval, and monitoring as part of our due diligence for governance and security.

Ultimately, Work IQ adheres to all of the security and governance policies and procedures in our tenant, preserving the trust that our security-first approach creates and maintains.

Key takeaways

The following are important considerations for data governance and security when you consider adopting Work IQ for your organization:

• With Work IQ, governance and security are top-line priorities. We made sure that Work IQ would always inherit the same compliance, access controls, and data protection policies that already govern Microsoft 365 data.
• Work IQ doesn’t introduce new data access—it changes how existing access functions. By packaging tenant data into a single intelligence layer, it facilitates easier agent builder access to the data you already have in your Microsoft 365 tenant.
• The distinction between discoverable and extractable data is central to safe AI deployment. Copilot and other agents can understand how work information is connected and referenced without exposing protected content beyond existing permissions.
• IT admins and leaders should consider the ramifications to their tenant. Work IQ makes agents more powerful and context-aware by opening up access to vast quantities of Microsoft 365 data, but IT professionals should always think through downstream effects on data security and governance.
• Work IQ surfaces governance gaps instead of masking them. When issues arise—such as misapplied sensitivity labels—the solution is not to restrict intelligence, but to strengthen data governance at the source.

Learn more

How we did it at Microsoft

• See how we’re governing AI agents at scale internally. This readiness guide explains how we approach security and compliance for AI agents at Microsoft.
• Read about our experience tackling Microsoft 365 Copilot governance. This guide details our internal governance efforts for Microsoft 365 Copilot as we rolled the tool out across our organization.
• Explore how we’re powering data governance at Microsoft with Purview Unified Catalog. This story explains how Purview is helping us redefine our internal approach to data governance.

Further guidance for you

• Learn more about governance and security for AI agents and find out how to keep your AI initiatives secure.

Chapter 3: How our employees experience Work IQ day to day

Transforming the way work happens at Microsoft

To understand how Work IQ shows up and impacts the workflows of people across our organization, we spoke to several Microsoft employees. They explained how Work IQ makes a difference in the results they’re getting from Copilot and other agentic AI tools and how the intelligence is supercharging their work.

Work IQ in Outlook    

Outlook email and calendars are the space where many of our employees feel the heaviest cognitive load of their day‑to‑day work. It’s also where Work IQ is surfacing some of the most innovative ways to help employees accomplish more.

“You open your Outlook in the morning and Copilot—by drawing on Work IQ context and through features like priority scoring and summarization—can help you see which messages need your attention first.”

Matthew Marzynski, principal product manager, core experiences, Microsoft Digital

Rather than treating messages and meetings as isolated items, Work IQ allows Copilot in Outlook to reason across email signals, conversation history, meeting patterns, and calendar behavior to deliver responses that reflect how work actually unfolds.

This means Copilot goes beyond keywords or unread status indicators to determine importance. Through Work IQ, it understands the context of each conversation—which threads are more urgent and relevant to your work and which are less vital.

“You open your Outlook in the morning and Copilot—by drawing on Work IQ context and through features like priority scoring and summarization—can help you see which messages need your attention first,” says Matthew Marzynski, a principal product manager for core experiences in Microsoft Digital. “Copilot is now beginning to offer proactive nudges to help you stay on top of what matters, surfacing what’s changed and what you need to focus on.”

The deeper context also aids Outlook in generating rich summaries of lengthy threads, which can highlight owners, decisions made, and next steps. This allows employees who are added to the thread or who have been away to quickly catch up on complex conversations without manually digging through seemingly endless past messages or related documents.

Marzynski frames Work IQ as an invisible intelligence layer that quietly reshapes how Outlook behaves over time. His core thesis is simple: Users never have to think about Work IQ; they just observe that Outlook is more helpful than before, and that their work gets easier.

“There are no complex commands to learn or rules to create. The intelligence works behind the scenes as you use Outlook,” he says. “Your inbox just gradually feels more relevant. Outlook adapts to how you work, rather than the reverse, and becomes more like an assistant instead of a filing cabinet of communications.”

Work IQ in Teams + Researcher Agent

Another immediate and tangible way our employees experience Work IQ is in Microsoft Teams meetings. The value begins the moment a meeting is recorded. Transcripts, speaker contributions, shared content, and AI‑generated summaries are automatically captured and folded into the attendees’ ongoing work context—without requiring manual note‑taking or follow‑up documentation.

Ray Peer is a senior product manager in Microsoft Digital who observed the power of Work IQ in a recent project he completed with our internal legal team. According to Peer, the team was struggling to find specific content in their data lake, which contains tens of thousands of documents, articles, and other content items.

“Based just on what people shared in that meeting, and what it knows about their work and about SIPOC diagrams, Researcher was able to generate a fully formed, detailed solution for me. That’s the intelligence layer at work.”

Ray Peer, senior product manager, Microsoft Digital

So, he facilitated a Teams meeting for a free-form process‑mapping discussion with a few members of Microsoft Legal. Days later, he put the meeting transcript into the Copilot Researcher agent and asked it to generate a structured SIPOC (Suppliers, Inputs, Process, Outputs, Customers) diagram and accompanying documentation.

He was amazed by the results.

“Based just on what people shared in that meeting, and what it knows about their work and about SIPOC diagrams, Researcher was able to generate a fully formed, detailed solution for me,” Peer says. “That’s the intelligence layer at work. It reasoned over what we said—there were no visuals shared or anything—and it came up with something that I could cut and paste into the final format. I used to have to do that manually, and it took hours.”

Work IQ connected the meeting transcript to the people involved, the SharePoint sites they used, and similar work done elsewhere in the organization. Copilot was able reason across different tools and unstructured data, rather than just treating the meeting transcript as a static artifact.

Note that this works differently from third‑party meeting tools, because the data never leaves the tenant. Work IQ treats Teams meetings as part of a continuous Microsoft 365 workstream—honoring permissions and sensitivity labels throughout—so conversations can become durable inputs for future work without adding risk or effort for employees.

Work IQ in SharePoint

In SharePoint, Work IQ is helping employees create, organize, and publish content by drawing on the rich context of their Microsoft 365 data. Rather than starting from a blank page or text block, content development is sped up as Copilot draws on their relationships, collaboration history, and metadata to help produce sites and documents.

“Copilot will recommend text changes, but also layout suggestions, image and graphic options, and other helpful assistance. It makes it easy to create more compelling content, more rapidly.”

Sam Crewdson, principal product manager, Microsoft Digital

For example, when you ask Copilot to create a new section in a SharePoint site—such as a project overview, status update, or other material—Work IQ enables the tool to look deeper than the prompt itself. When generating the content, it can draw on documents you’ve recently edited, your emails and Teams conversations, and related work happening across the organization. The output you get from Copilot is highly relevant and grounded in real work.

Sam Crewdson is a principal product manager at Microsoft Digital who has been a part of the SharePoint team for more than two decades. He’s excited about what Work IQ is enabling users to accomplish in the product using Copilot, as well as other agentic tools like Knowledge Agent (a domain-specific agent that can drill down on SharePoint sites and libraries).

“Copilot in SharePoint is now able to not only help you produce better written content, it’ll also offer more contextual and visual help,” Crewdson says. “Copilot will recommend text changes, but also layout suggestions, image and graphic options, and other helpful assistance. It makes it easy to create more compelling content, more rapidly.”

Another emerging scenario Crewdson described is conversational agentic authoring in SharePoint. In these workflows, employees refine their SharePoint pages by interacting directly with an agent—asking it to add sections, adjust tone, or suggest visuals. Over time, these agents will reduce repetitive setup steps and help teams move from draft to publish faster.

Across these experiences, Work IQ is helping shift SharePoint from a manual content creation tool to an application where agents automate everyday content tasks based on your overall work context and related Microsoft 365 data.

Key takeaways

Here are some things to remember when thinking about how Work IQ can impact your employee workflows:

• Work IQ reduces cognitive load in Outlook by understanding work context. By recognizing decision‑driven threads, collaboration patterns, and urgency over time, Copilot helps employees focus on what truly needs their attention without relying on manual rules or keyword searching.
• Email and calendar intelligence improves prioritization, summaries, and follow‑through. Work IQ allows Copilot to highlight owners, decisions, and next steps in long threads and nudge users toward timely action, based on how they typically work with colleagues.
• Teams meetings become durable inputs for future work when powered by Work IQ. Copilot and the Researcher agent can reason across meeting content, people, and related SharePoint work—creating structured outputs while honoring tenant security and permissions.
• Work IQ helps Copilot speed up and enrich content creation in SharePoint. By drawing on Microsoft 365 data, Copilot can generate more relevant content for your SharePoint sites and offer helpful layout and graphics suggestions that accelerate the site development process.

Learn more

How we did it at Microsoft

• See how to pick the right Copilot for the job. This post explores different Copilot agents and modules and discusses how to choose the correct one for your task.
• Discover how we deployed Copilot internally at Microsoft. This step-by-step guide breaks down the deployment and adoption process we followed to roll out this groundbreaking technology across our entire organization.
• Find out how we’re using AI to redefine work in Outlook. This blog post explores the latest AI features that our employees are using to work smarter in Microsoft Outlook.
• Learn five ways we’re getting more out of Teams in the era of Copilot and AI. This article details different Teams features that take advantage of the latest in AI technology.
• Read about the lessons learned using SharePoint agents at Microsoft. This post covers some of the insights gleaned through our deployment of different SharePoint agents.

Further guidance for you

• Check out the Microsoft Copilot Success Kit and learn how to accelerate your Copilot implementation with templates, onboarding guides, and technical readiness resources.
• Explore the company’s essential guide to Copilot adoption for practical steps to onboard, engage, and measure impact in your organization.

Chapter 4: Work IQ beyond Microsoft 365

Integrating Work IQ across the enterprise

As organizations adopt Copilot and other AI agents at scale, the question arises: How does Work IQ show up in different contexts? Is it something that only impacts your work in Microsoft 365 applications, or does it also play a role in external applications and other areas of your IT enterprise?

Based on our experience here at Microsoft, the answer is that Work IQ shows up differently depending on where it’s consumed, and those differences matter for admins, agent developers, and other IT professionals.

For most of our employees, Work IQ operates entirely behind the scenes inside Microsoft 365. It is not something users configure, enable, or interact with directly. By reasoning over your entire Microsoft 365 data graph, Work IQ improves the results that Copilot generates in apps like Outlook, Teams, Word, SharePoint, Copilot Chat, and GitHub Copilot.

In this mode, Work IQ is:

Built directly into the Microsoft 365 intelligence stack

Governed by existing permissions, sensitivity labels, and compliance policies

Invisible to users, with no new controls or UI to manage

Continuously learning from signals across emails, meetings, and files

You don’t “implement” Work IQ—it’s already present in first-party Microsoft products by default. If you have enabled Copilot, you are getting the benefits of Work IQ across all of these applications. 

Similarly, any agents you build for Microsoft 365 apps (such as using Agent Builder in Microsoft 365 Copilot) are scoped for use specifically in these apps, rather than outside of them. These agents do not require separate connectors, such as APIs or Model Context Protocol (MCP) servers, to access Work IQ. In fact, Work IQ MCP is a great tool to make your context ubiquitous to whichever agentic experience can be imagined.

Extending Work IQ beyond Microsoft 365: explicit by design

Implementation works somewhat differently outside of native Microsoft 365 experiences. When it comes to custom agents, line‑of‑business applications, or Azure‑hosted solutions, Work IQ does not show up automatically. In these contexts, it is intentionally enabled by our builders and governed by our administrators.

In these scenarios:

• Developers access Work IQ through APIs or MCP servers
• Admins explicitly control which capabilities are enabled or disabled
• Work IQ provides rich enterprise context without duplicating data
• Permissions and governance remain enforced at the tenant level

For us, this design is deliberate and has advantages. Rather than asking our developers to configure dozens of individual connectors for mail, calendars, files, and meetings, Work IQ offers them a single-entry point for enterprise context. Builder tools like Microsoft Foundry and Copilot Studio allow our teams to take the same Work IQ intelligence that Copilot uses and apply it to workflows that live outside Microsoft 365. Examples include automating newsletters, generating insights for account teams, or powering custom agents to handle specific scenarios.

The key distinction is accountability. Inside Microsoft 365, Work IQ is ambient. Outside it, Work IQ is a conscious architectural choice, one that requires actions upfront to enable.

Work IQ does not “open up new data” when used externally. It ports intelligence, not raw access, applying the same rules no matter where it’s consumed. At the same time, it gives organizations flexibility to decide when and how far that intelligence should travel.

This continuum—from implicit use inside Microsoft 365 to explicit use beyond it—also clarifies our roles:

• Our end users benefit without needing to learn anything new
• Our IT teams retain centralized control at the tenant level
• Our builders gain a faster path to context‑aware solutions

Work IQ works best when treated as a shared intelligence foundation, not a feature toggle. It is present by default where trust is already established, and it can be incorporated deliberately where your organizational requirements or innovation needs demand more reach.

Model Context Protocol servers and Work IQ

For organizations that move beyond native Microsoft 365 experiences and begin building custom agents, Model Context Protocol (MCP) servers are the primary mechanism for connecting those agents to Work IQ. While Work IQ is always available inside Copilot, MCP servers are what make much of that same intelligence accessible to agent builders.

At a high level, MCP servers are an open-standard technology (not proprietary to Microsoft) that act as governed tool interfaces to enterprise context. Each Work IQ MCP server represents a scoped slice of Microsoft 365 signals—such as email, calendar, Teams activity, or SharePoint content—and exposes them in a form that agents can reason over. Rather than wiring individual connectors or APIs for each workload, builders can rely on MCP servers to assemble relevant context automatically, while still honoring permissions, sensitivity labels, and tenant policies.

When we’re building agents, Work IQ becomes explicit, and MCP servers are how our builders declare their intent. This includes determining which types of enterprise context the agent needs, how broadly it should reason across work signals, and where governance boundaries apply.

From an IT perspective, MCP servers also provide a critical control point. Our administrators decide which Work IQ MCP servers are enabled in the tenant and which of our builders are allowed to use them. This ensures that extending intelligence beyond Microsoft 365 remains a deliberate choice rather than an accidental one.

Using these servers to connect with your enterprise data also represents real—but manageable—risk. They make existing permissions more actionable, which can amplify the impact of overshared content or weak data hygiene. The best practice is to treat these servers as governed infrastructure: enable them selectively at the tenant level, start with the minimum set required for defined agent scenarios, restrict usage to approved builders, and pair expansion with regular permission reviews and labeling discipline.

Your readiness plan should be to ensure that governance is in place, then selectively enable MCP servers where agents require deeper context. The servers are the bridge that lets agent builders tap into Work IQ safely, allowing you to bring enterprise intelligence into custom solutions without breaking the trust model that makes Copilot effective at scale.

Key takeaways

Here are some things to remember when thinking about how Work IQ shows up across your organization—especially if you plan to extend this intelligence into custom agents and applications:

• Work IQ is foundational inside Microsoft 365 and intentional outside it. Within Copilot experiences, Work IQ operates implicitly, while custom agents introduce a conscious decision to consume that intelligence through MCP servers.
• Governance principles don’t change when extending Work IQ, but they become more visible. MCP servers enforce existing permissions, labels, and tenant policies, making it critical that governance foundations are solid before agents rely on deeper context.
• Agent builders declare intent through MCP server selection. Choosing which Work IQ MCP servers to use defines what enterprise signals an agent can reason over and how broadly it reflects real work patterns.
• Preparing to extend Work IQ beyond Microsoft 365 is about readiness. Organizations that are already ready for Copilot can selectively enable MCP servers to unlock richer agent scenarios without introducing new security or compliance risk.

Learn more

How we did it at Microsoft

• Learn how we’re shaping AI management at Microsoft with Agent 365. This post discusses how we use the Agent 365 control plane to govern Microsoft 365 Copilot and agents internally.
• Read our complete guide to deploying AI agents internally at Microsoft. This Inside Track readiness guide covers all aspects of how we’ve implemented agentic AI across our organization.
• Discover how we’re unlocking enterprise AI extensibility with Copilot Studio. This story explores how we govern more advanced agents built with Copilot Studio, including connectors, orchestration, and scaled governance.
• Check out how we’re protecting AI conversations at Microsoft with Model Context Protocol security and governance. This article delves into MCP governance through secure-by-default architecture, automation, and inventory for faster, safer agent development.

Further guidance for you

• Read this Work IQ MCP overview on the Microsoft Learn site to find out more about how agents connect using Model Context Protocol servers.
• See the Work IQ Copilot reference on the Agent 365 Microsoft Learn site for further technical detail.

Chapter 5: Working with Work IQ: The Customer Zero impact

Change management lessons from our experience with an ambient intelligence layer

Work IQ wasn’t rolled out across our organization as an abstract platform decision or deployment milestone. Its development has been one aspect of our overall transformation into an AI-first Frontier Firm.

Along the way, Work IQ has been shaped by our long‑standing Customer Zero mission at Microsoft Digital: Using our own products at enterprise scale first, learning directly from how employees experienced it, and allowing those lessons to shape how the technology is refined and extended to customers.

In our tenant, Work IQ benefits emerged gradually through incremental improvements to relevance, context, and intelligence across Microsoft 365. These gains were driven by advances in AI that made it possible to interpret everyday work signals more effectively.

There was no formal product implementation or adoption campaign when we launched Work IQ at Microsoft. As ambient infrastructure, Work IQ is an unseen part of all employee workstreams—nearly every experience benefits from it. At the same time, the power of Work IQ depends on everyone in our organization being effective stewards of their own unstructured data, preserving security, governance, and relevance.

Enablement and adoption

To fully realize the value of Work IQ, we have found that organizations must invest in the foundational behaviors that make their organizational knowledge accessible. One of the key steps in this effort is enabling and encouraging the use of meeting transcripts. Work IQ depends on the artifacts of daily work to build context, and without transcripts, a significant portion of meeting insights and decisions remain inaccessible to the intelligence layer.

Making transcription a standard part of our employees’ everyday collaboration proved essential. Transcripts create a durable, searchable record that Work IQ can connect to documents and actions, helping employees quickly understand what happened, even if they weren’t present. When paired with existing governance controls like sensitivity and meeting labels, organizations can capture this data securely while unlocking great value from this collective knowledge.

This is actually a cultural shift.

We gave our teams clear guidance and encouraged meeting transcription as part of their normal workflow. When paired with the enhancements to meeting recaps in Microsoft Teams, this becomes a powerful tool for preserving and leveraging organizational knowledge.

Of course, Copilot adoption and training efforts were also a vital part of our getting the most from Work IQ. Our employees needed demonstrations of all the things that Copilot could help them accomplish, along with encouragement to jump in and try it out for themselves. Our data shows that internal AI usage has grown significantly over time—from a few thousand users to hundreds of thousands across the company—in large part due to:

• Employee-driven champions programs
• Scenario‑based learning efforts
• Timely and consistent internal communications

Usage also grew internally as our product teams continually refined our AI tools, aided by our collection of user feedback on agentic answers to identify low-quality output and irrelevant detail.

Another major insight we captured was the importance of persistent memory to the Copilot and Work IQ experience. Through our work as Customer Zero, we collected a large volume of feedback from employees indicating that this was a priority—users should not have to repeatedly explain who they are or what they are working on.

The experience was subsequently improved, and Work IQ now helps enable Copilot to remember user history and tailor responses accordingly—delivering summaries for communicators and deeper technical detail for engineers, for example.

Our Customer Zero efforts also validated a critical governance principle for us. As intelligence improved, some teams were surprised by how much context Copilot could surface. In every case, investigation showed that the underlying data access already existed. Work IQ did not change permissions or expose new data—it made existing relationships more visible. This reinforced the importance of strong data hygiene, sensitivity labeling, and permission management as prerequisites for trusted intelligence.

Ultimately, our work as the company’s Customer Zero validated that Work IQ is best understood as shared infrastructure. Its value compounds when organizations focus on readiness—governance, learning, and trust—and allow intelligence to scale naturally across work, rather than treating it as a feature to deploy.

When these conditions are in place, Work IQ quietly raises the quality of Copilot and agent experiences without adding complexity for users or additional burden for IT.

Key takeaways

As you consider how Work IQ might take shape in your own organization, consider these observations from Microsoft Digital’s Customer Zero experiences with this new intelligence layer:

• Meeting transcription is the key. Making sure all meetings are transcribed is essential for Work IQ, so it can build context on how work happens in your organization. This is a technical and cultural change that you need to facilitate and encourage.
• Awareness and learning are keys to usage and feedback. Our internal Copilot adoption grew when employees were shown practical scenarios and encouraged to experiment, supported by champions programs and ongoing internal communication.
• Change management drives results. Use employee champions, role-based immersive learning, and timely internal communications to help your employees understand what Work IQ is and how it can help your enterprise maximize the value of AI.
• Treating Work IQ as shared infrastructure unlocks compound value. When governance, learning, and trust were in place, intelligence could reason across all our rich unstructured data —improving Copilot and agent experiences without adding additional work for users or IT.

Learn more

How we did it at Microsoft

• Read about how we serve as Customer Zero in an AI-powered world. This story describes how Microsoft Digital’s role as Customer Zero has grown and evolved as we move into the Frontier Firm era.
• How to pick the right Copilot for the job. This post explores different Copilot agents and modules and discusses how to choose the correct one for your task.
• Recapping our meetings with Microsoft Teams Premium. This Inside Track story explains how Intelligent Recaps have changed the game when it comes to getting the most out of meetings at Microsoft.

Further guidance for you

• Check out the Microsoft Copilot Success Kit and learn how to accelerate your Copilot implementation with templates, onboarding guides, and technical readiness resources.

Where we’re heading: Work IQ, Fabric IQ, and Foundry IQ

Combining different layers of intelligence to transform the workplace

While impactful on its own, Work IQ is just part of larger story of how we’re using the power of rich data and agentic AI to transform how we work at Microsoft.

“While Work IQ can access your Microsoft 365 data, Fabric IQ will connect to your organizational data, such as analytics. Foundry IQ can leverage both, plus other domain data, to help developers build powerful agentic solutions.”

Naveen Jangir, principal architect, Microsoft Digital

Work IQ is one layer. It allows our AI tools to reason over unstructured data so this powerful resource can be a part of our larger enterprise intelligence system. But it also includes two other aspects of this three-layer system—Fabric IQ and Foundry IQ. Combined, these three capabilities enable organizations to take full advantage of your knowledge estate to forge the AI-powered workplace of the future.

“While Work IQ can access your Microsoft 365 data, Fabric IQ will connect to your organizational data, such as analytics,” says Naveen Jangir, a principal architect in Microsoft Digital. “Foundry IQ can leverage both, plus other domain data, to help developers build powerful agentic solutions.”

Here’s how these capabilities work together in complementary roles to impact how work gets done at Microsoft:

• Work IQ handles unstructured data—like documents, emails, PDFs, and web content—by extracting meaning and context from human language.
• Fabric IQ operates over structured data—like tables, databases, metrics, events, and transactions—to bring consistency and analytic rigor to our work.
• Foundry IQ provides the knowledge-grounding layer, where entities, relationships, and ontologies allow reasoning to stay aligned with enterprise truth.

While each component is powerful on its own, the deeper value is what becomes possible when they are used together.

The intent is to enable agents that can reason across all enterprise knowledge, regardless of where it originated or how it was stored. An agent should be able to read a policy, connect it to operational data, understand who and what is involved, explain its conclusions, and take an action (if desired) through a shared ontology based on organizational context.

That kind of capability can’t emerge just from information retrieval. It requires shared meaning across systems, content, and data types.

This is where the role of Work IQ becomes especially important. We have found that unstructured data contains some of the most critical institutional knowledge an organization has, but it rarely arrives in a form that is ready to be reasoned over. Documents reference people, systems, processes, and timelines in ways that make sense to humans, but not to machines. They can also fall out of date or represent a draft state that was never meant to be presented as verified information.

Work IQ bridges this gap by transforming the raw text into structured understanding, without stripping away nuance.

“Work IQ is already helping us change the way that work gets done. Instead of hunting for information or stitching context together manually, our employees can focus on decisions, creativity, and outcomes—because the intelligence is already there, working with them every day. It’s an integral part of preparing our organization for our agentic AI future.”

Vijaya Alaparthi, principal group product manager, Microsoft Digital

The crucial mechanism for that transformation is entity extraction, paired with a shared ontology. When a document mentions an employee, a system, a regulation, or a product, Work IQ identifies that reference as something concrete and reusable. Over time, those entities become the connective tissue between unstructured content, structured records in Fabric IQ, and the semantic backbone that Foundry IQ relies on to ground reasoning in the agents we create.

We can already see signs of this promised future at Microsoft today. Take a tool like our Employee Self-Service Agent, which we launched late last year. What before was a collection of static HR documents becomes a living knowledge system: policies are parsed, roles and eligibility criteria are extracted, and guidance is grounded in an understanding of employee role and location. The agent can answer a question and explain why the answer applies, because it understands both the document and the organizational context behind it.

This is why Work IQ is such a strategic capability. Improving document quality, normalizing metadata, resolving entities, and establishing governance are not one-off hygiene tasks. They expand what future agents will be able to do safely and reliably. The more coherent your unstructured data becomes, the less guesswork agents must do and the more context they can absorb.

“Work IQ is already helping us change the way that work gets done,” says Vijaya Alaparthi, a principal group product manager in Microsoft Digital. “Instead of hunting for information or stitching context together manually, our employees can focus on decisions, creativity, and outcomes—because the intelligence is already there, working with them every day. It’s an integral part of preparing our organization for our agentic AI future.”

For us, the direction forward is clear. The better your data foundation, the more capable—and trustworthy—your agents become. As unstructured and structured knowledge converges, intelligence stops being a set of isolated features and becomes a system.

Organizations that invest in technology like Work IQ to harness their unstructured data as enterprise knowledge are the ones that will deploy the most capable agents going forward and will be best positioned to take advantage of the agentic future.

Key takeaways

If you want your organization to be able to use Work IQ to propel your own agentic transformation, consider what we’ve learned on our journey:

• Work IQ transforms unstructured enterprise data into actionable intelligence. By reasoning over emails, documents, meetings, and chats, it unlocks institutional knowledge that was previously fragmented and underused.
• The intelligence operates as foundational infrastructure, not a user-facing feature. Work IQ runs continuously behind the scenes across Microsoft 365, improving Copilot and agent responses wherever they appear without configuration.
• Context is what makes Copilot feel truly intelligent. By combining signals from collaboration patterns, conversations, documents, and more, Work IQ enables agents to respond based on how work actually happens, not just what information can be retrieved.
• Security and governance remain intact because Work IQ inherits existing controls. It doesn’t create new access to data; it reveals relationships while fully honoring permissions, sensitivity labels, and compliance policies.
• Employees experience Work IQ as reduced cognitive load, not added complexity. Inbox relevance, richer summaries, and clearer follow-through improve naturally over time.
• Using Work IQ beyond Microsoft 365 is a deliberate, governed choice. MCP servers allow builders to bring enterprise context into custom agents while giving IT teams clear control over scope, access, and risk.
• Work IQ is the foundation for the next generation of agentic intelligence, especially when combined with Fabric IQ and Foundry IQ. The more coherent and well-governed your unstructured data is today, the more capable, explainable, and trustworthy your future agents will become.

Learn more

• Explore how Work IQ is helping supercharge AI usage at Microsoft.
• Discover our chapter-by-chapter internal deployment guide for Microsoft 365 Copilot.
• Check out how we’re embracing this agentic moment at Microsoft.
• Dive into a Microsoft Learn overview of different Work IQ agent tools.
• Read an IT playbook for the AI era and learn how we’re becoming a Frontier Firm at Microsoft.
• See how we’re using Model Context Protocol security and governance to protect AI conversations at Microsoft.
• Learn about our experience governing AI agents at scale in this readiness guide.
• Find out how we’re unlocking enterprise AI extensibility with Copilot Studio.

Try it out

Get a closer look at Work IQ.

The post Intelligence on tap: How Work IQ enables AI and agents at Microsoft appeared first on Inside Track Blog.

In an always-on business environment, even routine collaboration can be overwhelming. But in an environment of Frontier Transformation, this challenge represents an opportunity for AI.

Our employees don’t need to handle all their work alone anymore, because agents can now extend their responsiveness and reach. Here in Microsoft Digital, the company’s IT organization, one of those AI agents is acting as a digital twin for just that purpose. It’s called DigitalMe, a personal virtual proxy designed to keep work moving when our employees are busy with other tasks.

Always-on knowledge without always-on employees

Large meetings generate a constant stream of questions, side conversations, and follow-up items. They’re often more than a single presenter or moderator can manage in real time. Important insights get buried in chat threads, queries go unanswered, and valuable momentum gets lost.

For our teams at Microsoft, this challenge became especially visible during large-scale readiness sessions, where subject matter experts found themselves inundated with requests for clarification and guidance.

“In order for our transformation into a Frontier Firm to be successful, we need to step back and ask what works well for employees, what doesn’t work well, and where agents can help.”

Stephan Kerametlian, senior director, Microsoft Digital

That’s not the only place where employees can use an extra hand. When people are out of the office, that doesn’t mean work stops. Their coworkers often need access to their colleagues’ knowledge to move mission-critical work forward, even when they’re not reachable.

“In order for our transformation into a Frontier Firm to be successful, we need to step back and ask what works well for employees, what doesn’t work well, and where agents can help,” says Stephan Kerametlian, a senior director in Microsoft Digital. “We’re crossing the horizon into human-led, agent-operated patterns of work.”

One team in Microsoft Digital created DigitalMe to explore what that future could look like in practice.

Go deeper

Becoming a Frontier Firm: Our IT playbook for the AI era

DigitalMe: A personal digital twin for Microsoft employees

For the members of our Employee Experience Success team responsible for adoption efforts around Microsoft 365 Copilot and Microsoft Copilot Studio in the Greater China Region, readiness meetings were becoming unwieldy because of attendee questions.

“Our purpose was to use as little code and as much natural language as possible so people could modify their own personal DigitalMe easily. In Copilot Studio, you can manage agents as a solution. So users can just download and import a zip file, modify an agent like DigitalMe according to their business context and preferences, then use it.”

Ju Bu, business program manager, Microsoft Digital

The team wanted a way to focus on running the meeting while simultaneously providing their knowledge to participants. They decided to create an agent to help deal with the deluge of queries: DigitalMe.

At its core, DigitalMe is a personal, context-aware digital twin with versions that operate in both Microsoft Teams and Microsoft Outlook. It draws on the same knowledge bases and resources that its user can access, for example, SharePoint sites and Teams channels.

The team designed DigitalMe in Microsoft Copilot Studio and prioritized a low-code approach. At most, the creators used code to build 15–20% of the agent and accomplished the rest using natural language prompts.

“Our purpose was to use as little code and as much natural language as possible so people could modify their own personal DigitalMe easily,” says Ju Bu, a business program manager in Microsoft Digital. “In Copilot Studio, you can manage agents as a solution. So users can just download and import a zip file, modify an agent like DigitalMe according to their business context and preferences, then use it.”

Equipped with an employee’s full knowledge base, DigitalMe can respond in Outlook and Teams on its human counterpart’s behalf. To ensure transparency, a label appears at the beginning of each message indicating that it originates from the agent.

DigitalMe also reinforces context for the requester by including their original question in quotations. Finally, the agent @-mentions the recipient to notify them effectively.

The team identified two primary use cases for the agent:

• Moderating live sessions. In large meetings, DigitalMe acts as an always-on co-moderator, answering questions in real time using scoped, preloaded knowledge. By speaking for them in the meeting chat, it helps presenters stay focused while ensuring attendees receive timely, accurate responses. Surfacing information instantly enhances both the efficiency and quality of the session. DigitalMe has the added advantage of being able to pull from resources the presenter might not recall in the moment. Over time, the agent captures and reuses questions and answers, turning live engagement into a growing knowledge base.
• Extending employee availability. DigitalMe also provides a way for employees to remain responsive when they’re out of the office. It can monitor Teams chats or incoming emails, generate context-aware replies, and surface relevant knowledge for colleagues without human intervention. In practice, it’s proven especially valuable for teams distributed across widely different time zones and for handling project handoffs during onboarding or time-off scenarios.

A key advantage of DigitalMe is its ability to move beyond simple question-and-answer use cases. In some scenarios, it can also trigger workflows like creating tasks or capturing frequently asked questions.

“Our vision was that DigitalMe shouldn’t just be an assistant. It should function as our digital twin in the cyber world.”

Kai Cheng, program manager, Microsoft Digital

It was important to incorporate human-in-the-loop capabilities. When DigitalMe encounters gaps in its knowledge, it can flag those moments for follow-up, prompting users to refine and expand their knowledge sources. It represents another way that human-led, agent-operated processes continuously improve outcomes.

“Our vision was that DigitalMe shouldn’t just be an assistant,” says Kai Cheng, a program manager working in change management, digital transformation, and AI in Microsoft Digital. “It should function as our digital twin in the cyber world.”

In live sessions, DigitalMe has helped presenters stay focused while maintaining high levels of engagement, responsiveness, and support for participants. Employees are increasingly using it to bridge time zones, support knowledge transfer, and keep projects moving in their absence.

Extending the impact of DigitalMe

After seeing DigitalMe’s early success, our global readiness and adoption professionals identified the agent as an opportunity to turn individual innovation into a scalable capability. After templatizing the agent in collaboration with its original creators, we’ve now included it in our Agent Starter Kit. This resource makes it easy for employees to create their own personal versions of several useful agents.

“Employees often think building an agent might be complex and time-consuming, and that limits their willingness to try and turn their ideas into working solutions. But tools like this show them how easy it can be.”

Alexandra Jones, director of business programs, Microsoft Digital

Our Agent Starter Kit walks employees through importing a ready-made agent, connecting it to their knowledge sources, and adapting it to their specific workflows. This approach has shifted DigitalMe from a single solution into a repeatable pattern, helping employees across the company move from curiosity to hands-on adoption. We’ve also incorporated the Agent Starter Kit into our Agent Launchpad skilling program to accelerate our employees’ agentic expertise as part of a Frontier firm

There’s an added benefit as well. By getting tools like DigitalMe into people’s hands through templatized versions they can modify and configure themselves, we’re highlighting how easy it can be for even nontechnical workers to build agents themselves.

“Employees often think building an agent might be complex and time-consuming, and that limits their willingness to try and turn their ideas into working solutions,” says Alexandra Jones, director of business programs in Microsoft Digital. “But tools like this show them how easy it can be.”

For organizations that want to replicate this kind of solution, the path is increasingly straightforward. By lowering the barrier to entry with templatized agents and no-code tools, our team in Microsoft Digital has demonstrated that any employee can build tailored, high-impact assistants without deep technical expertise.

“The goal of Frontier Transformation is that AI is just there as you’re working, helping you practically do your job to enhance the experience and add value in real time.”

Kevin Wooldridge, senior director of digital transformation, Microsoft Digital

Looking ahead, we’re exploring ways to deepen these capabilities by adding memory and behavioral context so DigitalMe can better reflect individual working styles. The goal is to evolve it from a helpful assistant into a more complete digital representative.

Together, these advances point toward a future where employees routinely work alongside agents that grow, learn, and contribute more over time.

“DigitalMe is an example of the genuine, practical application of agentic use in the flow of work,” says Kevin Wooldridge, senior director of digital transformation in Microsoft Digital. “The goal of Frontier Transformation is that AI is just there as you’re working, helping you practically do your job to enhance the experience and add value in real time.”

The post Meet DigitalMe: Our AI digital twin that works on our behalf appeared first on Inside Track Blog.

“The pace of AI innovation has far outstripped how people learn. The old model of static catalogs, fragmented experiences, and a mindset of ‘consume content and move on’ doesn’t work in this new world, where roles are evolving in real time and every employee is expected to be AI proficient.”

Kavitha Radhakrishnan, general manager, Global Skilling

At Microsoft, this shift exposed a critical gap for us. While access to learning content has expanded dramatically, clarity about how to build and maintain skills has not kept pace. Our IT professionals often know they need to build AI capabilities but struggle with where to start, how to prioritize, and how to align their growth with business outcomes.

This is where our Global Skilling team identified an opportunity.

“We started with a simple observation: The pace of AI innovation has far outstripped how people learn,” says Kavitha Radhakrishnan, a general manager in Global Skilling product development. “The old model of static catalogs, fragmented experiences, and a mindset of ‘consume content and move on’ doesn’t work in this new world, where roles are evolving in real time and every employee is expected to be AI proficient.”

This situation led to the development of a cutting-edge solution: The AI Skills Navigator.

Try it at your company

Start your journey with our AI Skills Navigator

From content overload to guided capability building

At its core, AI Skills Navigator represents a shift in how learning is designed. Instead of asking learners to navigate sprawling catalogs of courses, the platform is built to guide them through a purposeful journey tied to their role, their goals, and the demands of their organization.

“Traditional learning catalogs answer the question, ‘What can I learn?’” Radhakrishnan says. “AI Skills Navigator answers the question, ‘What do I need to learn next—and why does it matter?’”

For IT professionals, that difference is significant:

• Learning paths are aligned to real-world scenarios and roles
• Content is curated and structured rather than fragmented
• Progression moves from foundational understanding to applied capability
• Skills are validated through credentials that signal actual proficiency

This approach helps IT teams move beyond passive learning and toward what Microsoft describes as “active capability building at scale.”

How AI Skills Navigator works for IT professionals

AI Skills Navigator is designed to meet IT professionals where they are, whether they are building foundational understanding, creating agents, or deploying and managing AI-powered solutions in production.

The experience is anchored in four key principles:

1. Curated skilling playlists aligned to real roles. Learners engage with curated playlists mapped to their role and responsibilities. These playlists guide progression from foundational proficiency to deep expertise and leadership with AI.
2. Applied skills, not just content consumption. The platform emphasizes hands-on, lab-based experiences where learners build and demonstrate real capabilities. Applied Skills credentials validate what learners can do, not just what they have completed.
3. Multimodal learning in the flow of work. Content is delivered in formats that fit how professionals learn day to day, including interactive modules, video, and audio-first experiences like podcasts. This makes it easier to build skills without stepping out of the workflow.
4. Skills validation with organizational visibility. Progress and credentials give individuals a way to demonstrate expertise. At the same time, organizations gain visibility into skill development and readiness at scale.

Behind the scenes, the experience is designed to deliver personalization at scale.

“The most important architectural decision we made was treating personalization as a ‘data and signals problem’ before it became a model problem,” says Iliyas Chawdhary, a principal group software engineering manager in the Global Skilling product group. “We built AI Skills Navigator on a modular foundation: a unified content catalog, separate skills and roles taxonomy, an identity and profile layer, and a recommendation surface connected through well-defined contracts. That separation enables us to make updates without rewriting the experience.”

By separating content, roles, identity, and recommendations into modular components, the platform can continuously evolve as technologies and job expectations change.

“The most consistent feedback from IT practitioners is that they need to move quickly from understanding AI to actually operating it.”

Priya Vaidyanathan, director of product management, Global Skilling

A differentiated approach to AI skilling

While many platforms provide access to AI learning content, AI Skills Navigator is differentiated by how it connects learning to real-world outcomes.

“The most consistent feedback from IT practitioners is that they need to move quickly from understanding AI to actually operating it,” says Priya Vaidyanathan, director of product management for Global Skilling. “The focus on governance, security, and how to enable their organizations without slowing innovation is a key differentiator for us.”

AI Skills Navigator is different from other learning experiences in several other ways:

• Built around roles and tasks, not course catalogs. Content is organized into curated playlists aligned to roles and real work scenarios. Learners are not choosing from a library of courses; they are guided to build the specific skills needed to perform in their role, from first exposure to applied execution.
• Orchestrated by specialized agents, not a single recommendation engine. Multiple agents work together to create playlists, guide learning sessions, and ensure content quality. This allows the experience to adapt to the learner, while maintaining grounding in trusted, curated Microsoft content. The result is guidance that is both personalized and reliable.
• Designed to build capability over time, not deliver one-time learning. The platform is designed for repeat engagement. As learners return, recommendations evolve based on progress, feedback, and emerging skills; this enables continuous skill development rather than a one-time completion model.
• Embedded into how work happens, not separate from it. Integration with Microsoft 365 Copilot brings skilling into the tools professionals already use and learning happens alongside real tasks, making it easier to apply skills immediately instead of learning in isolation.

Turning learning into team capability

For organizations, one of the most powerful features of AI Skills Navigator is the ability to align teams around shared learning goals. Skilling playlists enable leaders to define capability journeys that map directly to business priorities.

Instead of assigning generic training, leaders can create structured paths that guide teams toward specific outcomes, like becoming AI literate, managing agents in the enterprise, or building expertise in agent development. This approach transforms learning from an individual activity into a shared experience.

For IT professionals, this means learning is no longer abstract—it becomes directly connected to their role, their team, and the transformation initiatives they support.

Building momentum with an AI Skills Fest

To accelerate skills development through a moment of shared learning, we’re hosting our second global AI Skills Fest initiative in June 2026. The annual event is designed to bring focus, energy, and community to the AI Skills Navigator experience.

AI Skills Fest brings together:

• A global audience of learners across different roles and skill levels
• Curated learning experiences aligned to real-world scenarios
• Opportunities to engage, practice, and validate new skills

The initiative builds on the success of last year, when we brought together more than 126,000 participants in a single day of learning to achieve a Guinness World Record for AI skilling participation. This milestone demonstrated both the demand for AI skills and the power of creating a shared learning moment at global scale.

In 2026, the focus is shifting from the record itself to sustaining long-term engagement. Our AI Skills Fest is designed to help learners discover the right entry points into AI Skills Navigator and build momentum that continues well beyond the event.

“We want learners to think of it less as a single event and more as a catalyst for ongoing skilling at scale,” Radhakrishnan says.

Bringing AI skilling directly to Inside Track

To make these learning opportunities even easier to discover, we’re taking the next step by integrating AI Skills Navigator content directly into the Inside Track experience. This integration will provide IT professionals with:

• Direct access to curated learning journeys, aligned to Inside Track content
• Seamless pathways from insight to action
• A clearer connection between Microsoft’s own transformation story and the skills required to replicate it

For our readers, this creates a new kind of experience. Instead of simply learning how Microsoft approaches AI, you’ll be able to immediately start building the skills needed to apply those insights at your own organizations.

Look for curated deep links from Inside Track stories into AI Skills Navigator starting in the second half of 2026. Additionally, we’ll be adding links to our site navigation and Careers page to make it simpler to help you discover and build role-specific AI skills.

“From an assistant, to a coach, to a learning companion—the endpoint doesn’t feel like a learning platform at all. It just makes you better at your job.”

Iliyas Chawdhary, principal group software engineering manager, Global Skilling

A new model for learning in the era of AI

AI is reshaping how organizations operate, how teams collaborate, and how work gets done. For IT professionals, staying relevant means continuously building new capabilities.

AI Skills Navigator represents our answer to that challenge. The initiative moves beyond static content to create a guided, adaptive, and integrated learning experience. AI Skills Navigator just feels different—and better—than other learning platforms.

“From an assistant, to a coach, to a learning companion—the endpoint doesn’t feel like a learning platform at all,” Chawdhary says. “It just makes you better at your job.”

The era of AI demands a new approach to learning, and that approach is built on a foundation of role clarity, relevance, and continuous growth.

Key takeaways

If you are thinking about promoting AI skilling among your own employees, keep the following in mind:

• Start your journey now. Explore the IT Professional playlist in AI Skills Navigator to identify the skills most relevant to your role.
• Align learning to outcomes. Don’t just take courses: Define the AI capabilities your role or team needs, then use structured playlists to guide progress.
• Make learning continuous. Plan for regular, incremental skilling rather than one-time training events to keep pace with AI innovation. The AI Skills Navigator playlists are constantly being updated to help you keep up with the pace of change.
• Leverage AI-powered guidance. Use AI-driven recommendations, playlists, and coaching experiences to accelerate learning and reduce time to value.

Try it out

Visit our AI Skills Navigator site to start your learning journey.

Related links

• Learn how we’re skilling up for the future of work at Microsoft with Agent Launchpad.
• Discover how we enabled meaningful AI adoption internally with a Microsoft 365 Copilot Expo.
• See how we drove Microsoft 365 Copilot adoption with a virtual skilling event.
• Find out what it’s like to work in Microsoft Digital in the era of AI.
• Explore the new age of AI-led engineering in IT at Microsoft.

The post Building AI skills for the future: How we’re reimagining learning with AI Skills Navigator appeared first on Inside Track Blog.

That’s why, as part of our broader approach to AI at Microsoft, we—Microsoft Digital, the company’s IT organization—are building a framework to measure the impact of the AI investments we’re making on behalf of the company.

“If we want to measure the business impact of AI, the conversation quickly moves toward identifying the agents or AI efforts that are driving the most value and satisfying business outcomes. We know that those conversations can be complex, so we use a value measurement framework to capture and assess the signals we have available.”

Don Campbell, principal group technical program manager, Microsoft Digital

Our framework is helping us move from AI enthusiasm to AI accountability. It creates a common way for us to talk about value across our different initiatives, teams, and business processes. It also helps us ask a harder, more specific question every time we assess the impact of AI at Microsoft Digital: If AI saves time, reduces costs, improves quality, or lowers risk, what changes are we making to take advantage of that?

We don’t have the full answer yet—we’re still improving the way we measure. Some of our signals are instrumented, some rely on strong hypotheses, and some need better telemetry. But we’re not waiting for perfect results to start learning.

“If we want to measure the business impact of AI, the conversation quickly moves toward identifying the agents or AI efforts that are driving the most value and satisfying business outcomes,” says Don Campbell, a principal group technical program manager in Microsoft Digital. “We know that those conversations can be complex, so we use a value measurement framework to capture and assess the signals we have available.”

Building a framework for AI business value

AI value doesn’t show up the same way everywhere. One investment we make might help employees complete a task faster, while another might improve quality, reduce risk, increase coverage, or lower operational costs. Some of that value can be measured directly, while in other cases it starts as a hypothesis that needs to be tested. That range is why we needed a common framework instead of a single metric.

Our value measurement framework helps our Microsoft Digital teams answer three basic questions before and after they build:

• What kind of value do we expect this AI investment to create?
• How will we measure that value?
• What will we do with what we learn?

We organize the answers to those questions around six value areas:

Revenue impact: How an AI investment contributes to our business growth, sales activity, customer targeting, or deal velocity.

Productivity and efficiency: How AI helps our people complete tasks faster, increase throughput, optimize processes, or automate work.

Security and risk management: How AI helps us identify, prevent, or manage security vulnerabilities, risk exposure, or Responsible AI compliance.

Employee and customer experience: How AI improves satisfaction, engagement, or the quality of a product or service experience.

Quality improvement: How AI improves deliverables, accuracy, confidence in outputs, or process quality.

Cost savings: How AI reduces our operational cost, improves our resource allocation, or helps us avoid future cost.

Our framework doesn’t require every AI investment that we make to create value in all six areas. In fact, that rarely occurs. A support automation scenario might focus primarily on productivity, employee experience, and cost avoidance. A security scenario might involve risk reduction, vulnerability coverage, or the ability to address more issues than a team could handle manually. A process quality scenario might relate to measures that are specific to a particular workflow and be harder to roll up into a single number.

“Measurement sources vary based upon the type of AI initiative. The six value areas give us a framework for measurement, but we need to observe and collect information where measurement starts to become practical. Teams need to understand the processes affected, including how long work takes, how many resources are involved, and what the workflow looks like before and after AI.”

David Laves, director business programs, Microsoft Digital

The framework creates consistency in how we talk about value, while giving teams room to measure what actually matters for their scenario. Some measures roll up easily, including cost savings, time savings, and certain risk measures. Others are more specific to the process being improved. Those measures still matter because they help business owners understand whether AI is changing the work in a meaningful way.

“Measurement sources vary based upon the type of AI initiative,” says David Laves, a director of business programs in Microsoft Digital. “The six value areas give us a framework for measurement, but we need observe and collect information where measurement starts to become practical. Teams need to understand the processes affected, including how long work takes, how many resources are involved, and what the workflow looks like before and after AI.”

Our framework also helps us make better investment decisions. Before we commit to an AI scenario, we can map the opportunity to the value areas that matter most, estimate the value we think it can create, and decide what needs to be measured. After implementation, we can compare results against the baseline, review the data with the right business and AI owners, and adjust the work based on what we’re seeing.

That last step is critical. Our framework creates a way for us to create an operating rhythm around AI value. It helps us take a promising scenario and prove its worth (or lack thereof) by establishing expected value, evaluating what is being measured, and deciding what to change because of the results.

It’s a continual evolution of business value measurement that prioritizes progress over perfection, and it’s helping ensure that our AI approach stays grounded in the outcomes that are most meaningful to our business.

Turning measurement into an operating rhythm

A framework only matters if teams use it to make decisions. For us, that means moving measurement out of one-off reporting and into a regular management cadence. We track our highest-business-value AI initiatives by priority and business function, then review the KPIs that show whether those investments are creating the value expected.

Some KPIs roll up cleanly. Our cost savings, time savings, and certain risk measures can be summarized across initiatives and discussed at a leadership level. Other KPIs stay closer to the individual scenario because they’re tied to a specific workflow, process, or business outcome. We need both. Our rollup metrics help our leaders see broad progress, while the scenario-level metrics help our teams understand what’s changing inside their work.

“We actually create monthly targets and end-of-year targets for every top AI-enabled initiative,” Campbell says. “Then we basically reiterate every single month with our leadership team to look at the value we’re driving and have conversations about it.”

That monthly rhythm helps us proactively manage AI value. If a measure is trending positively, we look at what’s working and where else the pattern might apply. If a measure is off track, teams can dig into the supporting data, review the assumptions, and decide whether they need to adjust the solution, the measurement, or the operating process around it.

This process supercharges prioritization. Our team here in Microsoft Digital has a large set of AI opportunities, and not every idea can move at the same pace. By mapping initiatives to value areas, estimating expected impact, and tracking results over time, we can have a more grounded conversation about where to invest, where to scale, and where to keep learning.

That discipline becomes more important as AI moves deeper into business processes. We don’t want teams to measure only adoption or usage if the real goal is a better business outcome. Usage matters, but it doesn’t tell the whole story. A tool can be used often and still fail to improve the process it was meant to change.

Applying the framework: Global Support

Consider the following example of applying the framework from our Global Support team. This team is currently examining how AI can help automate specific pieces of the ticket management process.

“In Global Support, the processes that often matter most from a value perspective are the ones with high repetition. If a process runs thousands of times a month and can operate autonomously, without human input, that’s where AI can deliver meaningful, measurable impact.”

David Finney, principal program manager, Microsoft Digital

As part of this effort, we examined a support process that depended on manual follow-up. In this process, after a Global Support team member marks an issue as resolved, the team waits for the user to confirm that the ticket can be closed. If the user doesn’t respond, the agent must follow up once a day for up to three days. After the third attempt, the agent simply closes the ticket.

This user flow gave us a practical way to test the value framework. It has repetition, because it runs often; it has autonomy potential, because the steps are deterministic and rule-driven; and it has a clear time-savings opportunity, because a human agent spends time checking the ticket, writing the follow-up, and sending the message. It also has a measurable implementation effort, because the data exists in the ticket process but the solution still needs to integrate with ServiceNow.

“In Global Support, the processes that often matter most from a value perspective are the ones with high repetition,” says David Finney, a principal program manager in Microsoft Digital. “If a process runs thousands of times a month and can operate autonomously, without human input, that’s where AI can deliver meaningful, measurable impact.”

Finney estimated that about 5,000 tickets a month execute this process. Because each ticket can require up to three follow-ups, that can create up to 15,000 manual email follow-ups a month. At about three minutes per follow-up, that’s roughly 750 hours of productivity spent on one small piece of the process each month.

The framework helps us look at that work through both value and effort. On the value side, we can evaluate repetition, autonomy potential, and time savings. On the effort side, we can assess whether the data exists, how complex the solution is, whether engineering work or system integration is required, and how long implementation may take.

“We started to evolve our conversation to, ‘So what?’” Campbell says. “You saved money, you saved hours. What did you do with it? Where did the actual business outcome sit?”

Don Campbell, principal group technical program manager, Microsoft Digital

That structure matters because a high-value opportunity still needs the right implementation path. In this case, the process is part of the ticket workflow, so the needed data exists. The complexity comes from integrating the automated agent with ServiceNow so it can interact with the ticket, check whether the user responded, send follow-ups, and resolve the ticket according to the defined process.

Instead of trying to automate all of support at once, the team identifies specific subprocesses where AI has a clear role and the value can be measured. “It’s taking a sort of bite-sized approach to AI rather than trying to solve for everything in one big go,” Finney says.

That’s the kind of practical example the framework is designed to surface. It helps us find work that’s frequent enough to matter, structured enough for automation, and measurable enough to prove whether the AI investment changed the process.

Moving from savings to reinvestment

Measuring value starts the next conversation. If an AI investment saves time, reduces cost, increases coverage, or improves quality, we need to know what happens next. The number is relevant, but the business outcome is more important.

“We started to evolve our conversation to, ‘So what?’” Campbell says. “You saved money, you saved hours. What did you do with it? Where did the actual business outcome sit?”

That’s the harder part of AI value measurement. A team might use AI to reduce time spent on repetitive work, but the real value depends on how that recovered capacity gets used. In some cases, the reinvestment path is clear. A team can point to more programs delivered, more backlog reduced, more issues reviewed, or faster service delivery.

In other cases, the value is harder to trace. Some AI improvements return small amounts of time to many employees. Those minutes matter, but it’s difficult to prove exactly where each person reinvested them.

We’re careful when it comes to measuring ROI. We know our leaders will ask for it, and it belongs in the broader value conversation. But we don’t want ROI at the center of the story before we have the right cost model, telemetry, and approved data to support it.

For now, we’re focused on the operating discipline: Define expected value, baseline the current state, instrument the AI-enabled process, track results, review the data, and act on what we learn. That discipline is teaching us a number of practical lessons:

• Measurement needs to be built into the design of the AI investment, not added after launch.
• Teams need a baseline for the current process, so they can compare it with the AI-enabled process.
• Teams need to pick measures that fit the scenario.
• Data must have clear ownership, because uncertain data weakens the conversation with business owners and leaders.

Consistency matters as much as the metric itself. When our teams review value on a regular rhythm, they can see trends, test assumptions, and adjust the solution or the process around it. Some measures will be mature, and others will be directional. Some will need better instrumentation. The point is to keep improving the quality of the measurement while keeping the conversation focused on business value.

We’re continuing to build our value measurement muscle across Microsoft Digital. We’re not looking for one perfect formula for every AI investment. Instead, we’re creating a repeatable way to define value, measure it, review it, and use it to guide our next action.

As our AI investments and overall strategy mature, that framework helps us stay honest about what we know, clear about what we still need to learn, and focused on the outcomes that AI is designed to improve.

Key takeaways

Here are five actions you can take to help measure the impact of AI investments at your organization, based on what we’ve learned in our own efforts:

• Start with business outcomes. Define the business result you want first, so you can measure whether the AI investment creates real value.
• Choose metrics that fit the scenario. Select measurement areas that match the workflow, such as time saved, cost reduced, quality improved, or risk lowered.
• Establish a baseline before launch. Capture current performance before implementation, which will enable you to compare results and show what changed.
• Review results on a regular rhythm. Check performance consistently with the relevant stakeholders so that you can spot trends and adjust quickly.
• Reinvest gains intentionally. Use the time, savings, or capacity that AI generates to deliver clear value and ROI, instead of treating efficiency as the final goal.

Try it out

Get started with Microsoft Agent 365.

Related links

• Learn how a data council is powering our unified data strategy at Microsoft.
• Discover how a Center of Excellence is powering the technical veracity of AI at Microsoft.
• Explore how we’re reshaping Microsoft with continuous improvement and AI.
• Find out why Responsible AI matters and how we’re infusing it into our internal AI projects at Microsoft
• Read about the enterprise tools and solutions at Microsoft AI.
• Check out more details about Responsible AI at Microsoft.
• Learn more about the opportunity for AI in the Microsoft 2026 Work Trend Index Annual Report.

The post Measuring the impact of our AI investments in IT at Microsoft appeared first on Inside Track Blog.

Our Global Collection team in the Microsoft Treasury division makes sure payments are seamlessly executed in our fast-moving global enterprise environment. However, our case managers were often losing valuable time figuring out things like who the right contact was for a given customer, which issues were likely to be challenged by a customer, and where an exception should be routed next. This information was spread across systems or buried in handoffs.

To solve these challenges, our team built a human-led, AI agent-assisted support system to reduce preparation time and streamline their processes.

“Building the AI assistance wasn’t the hard part,” says Kathy Brustad, a director in the Global Treasury and Financial Services division at Microsoft. “The hard part was reimagining the collection experience with AI front and center, and bringing the underlying infrastructure up to speed to get it there.”

In this post, we explain how we did it so you can learn from our experience.

“We have over 1,000 collectors around the world who perform collections for Microsoft. They had multiple systems they had to go to in order to find out things like the totality of the customer’s invoice and what conversations a different team had with the customer. The information was fragmented.”

Kathy Brustad, director, Global Treasury and Financial Services

Stitching together information across systems

Our AI agent is focused on helping our case managers prioritize high-value work by:

• Predicting late payments and possible customer disputes
• Summarizing customer case interactions for use by case managers
• Routing customer emails to the right collections manager faster and with greater precision Automatically matching payments to invoices
• Automatically responding to customer inquiries

“We have over 1,000 collectors around the world who perform collections for Microsoft,” Brustad says. “They had multiple systems they had to go to in order to find out things like the totality of the customer’s invoice and what conversations a different team had with the customer. All of this information was fragmented. We didn’t have a single view of how much a customer owed us.”

We started by consolidating these dispersed tools and systems into an SAP and Microsoft Dynamics 365 environment, creating a single source of truth for all relevant customer, invoice, and payment data.

On that foundation, we layered on Microsoft’s IQ intelligence platform to infuse semantic understanding and business context. That standardized our workflows by simplifying templates and worklists to reduce complexity and put consistent global practices into place. Routine communications became fully automated.

We then applied AI to improve payment matching accuracy from 40% to 90%, generate customer response drafts, and intelligently route cases to reduce time-consuming back‑and‑forth.

Copilot assistance was embedded directly into the daily workflow of our case managers to reduce administrative load by providing inline knowledge suggestions, summarizing calls, and automatically drafting replies. With these standardized automated workflows, we could apply 98% of payments within 48 hours.

“In a nutshell, this is the collection story: We have various agents and models deployed to assist our human agents with all the activities they have to do, saving hundreds of thousands of hours that we spent on manually tracking things before.”

Kathy Brustad, director, Global Treasury and Financial Services

Moving faster on ‘act ready’ work

Deploying the agent was only the starting point. The harder work was helping our collection team change established ways of working. Brustad described the shift as learning to “run it in a different way,” moving from manual, fragmented preparation toward workflows where prioritization, context gathering, and routing were increasingly supported within the system.

To make that shift possible, the team introduced a change management work stream program and role-based training focused on real, day-to-day scenarios alongside the rollout. By anchoring the work in clear business pain points and showing tangible improvements, our team saw how the new approach made their work easier. Each morning, the agent prioritized each case manager’s workload according to urgency and past client behavior so case managers could immediately focus on the accounts that were the most pressing.

We reduced repetitive communications using automatically drafted responses and automated statements.

“In a nutshell, this is the collection story: We have various agents and models deployed to assist our human agents with all the activities they have to do, saving hundreds of thousands of hours that we spent on manually tracking things before,” Brustad says.

After deploying this system to our case managers, we saw measurable improvements in both productivity and speed, including:

• Hundreds of thousands of hours unlocked annually in order to do more human-led high-value work rather than routine administrative tasks
• 40% reduction in call preparation time
• 2X growth in automatic cash applications
• 2.5X acceleration of customer inquiry resolution time

Operationally, the team also saw up to 60% reduction in inquiry handling time through inline suggestions, summarized calls, and automatically drafted replies. To ensure these improvements were real and repeatable, we emphasized observability in our evaluation approach. Our team tracked dollars collected through collections and hours worked to create productivity metrics.

Data, trust, and good governance

When introducing AI systems or agents into finance workflows, leaders often ask two questions:

1. Can we trust the outputs?
2. Can we govern the process?

“The biggest takeaway is to know your own process very, very well. You need to understand where all the bottlenecks and pain points are. Start from there to design the new agent-enabled process instead of saying, ‘I’m going to just inject the agent into my existing process.’”

Kathy Brustad, director, Global Treasury and Financial Services

For us, trust came from getting the basics right in the form of right-sizing our enterprise data, standardizing our workflows, and establishing clear ownership for each part of the work. When we tested early and included frontline users throughout the process, outcomes improved.

“The biggest takeaway is to know your own process very, very well,” Brustad says. “You need to understand where all the bottlenecks and pain points are. Start from there to design the new agent-enabled process instead of saying, ‘I’m going to just inject the agent into my existing process.’”

Embed custom agent assistance directly into the moments where time disappears, such as prioritization, preparation, routing, and drafting so adoption feels natural and can be measured. You can prove impact with a small set of metrics like cycle time, throughput, dollars collected, and hours saved, and iterate from there.

Key takeaways

Modernizing collections is about fixing the fundamentals first, before you add AI into the mix. As you begin to streamline your own finance workflows, keep these lessons in mind:

• Fix fragmented workflows before adding intelligence: AI delivers the most value when it’s layered on top of standardized processes and a unified data foundation rather than disconnected systems and ad hoc handoffs.
• Embed assistance where time is actually lost: Copilot-style support works best when it shows up directly in prioritization, preparation, routing, and drafting to reduce friction without changing how people work.
• Focus AI on high‑ROI decisions, not just automation: Predicting late payments, flagging likely invoice disputes, and surfacing context can help teams spend time where it matters.
• Design around the practitioner’s day: When work arrives prioritized and prepped, case managers spend less time chasing context and more time resolving exceptions.
• Measure what matters to prove impact: Cycle time, dollars collected, throughput, and hours saved provide a clear, repeatable way to track productivity gains and cashflow velocity.
• Pair generative AI with strong governance: Trust comes from clear ownership, standardized workflows, quality data, and ongoing human oversight.

Editor’s notes:

• SAP is an enterprise finance system that many organizations use to manage invoices, payments, and financial records in a single, centralized platform.
• All metrics cited are based on Microsoft internal data gathered during the writing of this article. They’re best read as directional signals from that period, and they may change as systems, processes, and behaviors evolve. Microsoft makes no warranties, express, implied, or statutory.

Try it out

• Build and test an AI‑assisted workflow with Microsoft Copilot Studio to see how agents can help prioritize work, summarize context, and route cases more efficiently.
• Start a Dynamics 365 free trial.

Related links

• Discover the five lessons we learned on how to make AI stick for sellers during our Copilot rollout.
• See how we revamped and modernized our Microsoft Treasury infrastructure using Azure.
• Explore how we’re supercharging our support experience with Microsoft Dynamics 365 and AI.
• Learn how we’re transforming the marketing function at Microsoft with AI.
• Read about our role as Customer Zero in an AI-powered world.
• Check out how we’re measuring the impact of Microsoft 365 Copilot and AI internally.

The post Streamlining finance cash collection at Microsoft with AI appeared first on Inside Track Blog.

For years, our security groups operated without consistent, policy‑based guardrails. As a result, we couldn’t uniformly control guest access to sensitive resources or apply governance consistently across different group types.

Addressing this required a complex, coordinated effort by our team here in Microsoft Digital, the company’s IT organization, and the Microsoft Entra product team.

“Because IT security is our highest priority at Microsoft, we knew we needed a better approach to limiting access to groups within our tenant. And we realized that Microsoft Entra was a powerful in-house solution that represented our best path forward to solve for this challenge.”

David Johnson, principal product manager architect, Microsoft Digital

The result is a new approach to sensitivity labels across the organization that strengthens our security posture, which benefits Microsoft and our customers.

“Because IT security is our highest priority at Microsoft, we knew we needed a better approach to limiting access to groups within our tenant,” says David Johnson, a principal product manager architect in Microsoft Digital. “And we realized that Microsoft Entra was a powerful in-house solution that represented our best path forward to solve for this challenge.”

Go deeper

Learn how to assign sensitivity labels to Microsoft Entra security groups.

Closing the security gap

Sensitivity labels for Microsoft 365 groups are labels that govern join and access restrictions for membership and sharing. They have been a product feature since 2020. But sensitivity labels for security groups—labels that enforce rules about who can join a group—had no equivalent.

This meant that organizations that wanted to govern who could join a security group or determine if guests are permitted and how group membership is managed had to either lock down the group creation process entirely, or rely on reactive scanning after the fact.

“Security groups are a key piece of our efforts to secure sensitive resources,” says Mohit Bhargava, a principal product manager on the Microsoft Entra team, which manages the Entra family of identity and network access products. “We wanted to apply policies to protect who could be in security groups so that the sensitive resources in those groups would remain secure.”

“Whoever gets into an Azure security group can have access to all the resources associated with the Azure subscription. That’s a potential high-severity threat.”

Basanth Kakumani, software engineer II, Microsoft Digital

The security risk is real. If an unauthorized guest account ends up as a member of a security group that governs access to an Azure subscription, that guest gains access to every resource inside that subscription.

“Whoever gets into an Azure security group can have access to all the resources associated with the Azure subscription,” says Basanth Kakumani, a software engineer II in Microsoft Digital. “That’s a potential high-severity threat.”

Another priority was the need for consistency across experiences.

“Microsoft 365 groups have supported labeling for a very long time,” Bhargava says. “Customers have an expectation that there’s parity across group types, so that they can govern them uniformly. That was another driving factor for this work.”

Security groups reuse the same sensitivity labels already configured for Microsoft 365 groups and SharePoint sites in Microsoft Purview—so admins don’t need to create or manage a separate set of labels. This reuse reduces configuration overhead and supports a more consistent governance model across group types.

Security workarounds, and why they fell short

Without sensitivity label support, we had to make do with alternative solutions. The most common one was simply preventing certain users from creating any security groups at all.

In the Microsoft tenant, this meant that employees who needed a security group had to fill out a form that had custom business logic behind it.

“We had on-premises, Active Directory, synchronization, tooling, and customization,” Johnson says. “This caused latency, from the time you created your group to the time it would show cloud membership. If you wanted to manage your membership, you had to do it on premises, AD, and then wait for it to sync to Entra.”

Neither centralized control nor reactive governance was a satisfying solution to prevent policy violations.

“This is really about making reactive things more proactive. We want to catch problems before they occur.”

John Begley, principal software engineer, Microsoft Digital

Typically, IT is going to manage this in one of two ways: Either we turn off self-service and manage everything on behalf of users, or we do reactive governance, which includes scanning groups and looking for policy violations.

Those aren’t super effective at preempting violations.

“This is really about making reactive things more proactive,” says John Begley, a principal software engineer in Microsoft Digital. “We want to catch problems before they occur.”

A collaborative solution

Coming up with a solution to this challenge required a genuine partnership.

We at Microsoft Digital approached the Entra product team and explained the problem we were trying to solve. Rather than simply handling this as a feature request, the two teams agreed to a co-development arrangement.

“Having access to a very large customer who cares deeply about security was extremely helpful. If it works for Microsoft, which is so complicated and huge, it’s going to work for smaller-sized tenants too.”

Mohit Bhargava, principal product manager, Microsoft Entra

Microsoft Digital team members would work alongside Entra engineers as the feature was built, serving simultaneously as implementation partner, design critic, and test environment—what we like to call our Customer Zero role.

Bhargava found the partnership equally illuminating from the product side.

“Having access to a very large customer who cares deeply about security was extremely helpful,” he says. “If it works for Microsoft, which is so complicated and huge, it’s going to work for smaller-sized tenants too.”

For Begley and his team, working closely with the product team revealed how complex the solution actually was.

“Both the product team and Microsoft Digital walked into this thinking a fix was going to be simpler than what it turned out to be,” Begley says. “It’s been eye-opening to see how the product is built, how it runs, what all the moving parts are. We learned early on that there was significant co‑development happening within Entra itself, across teams with very different areas of expertise.”

That dynamic played out in specific feature decisions. The team’s original plan did not include support for agent access controls and didn’t include the ability to prevent AI agents from joining sensitive security groups. This is something the product group quickly addressed and resolved after our team in Microsoft Digital raised it as a concern.

“One of the first customers who raised it was Microsoft Digital,” Bhargava says. “They said we needed need to start thinking about it ahead of time to get ahead of the problem.”

Sensitivity labels for Microsoft Entra cloud security groups are now in public preview. The same labels you publish in Microsoft Purview for Microsoft 365 groups and sites now apply to Entra security groups. Visit Microsoft Learn for scope, supported scenarios, and current preview behaviors.

Changes afoot for IT admins and employees

The practical impact of this solution lands on both sides of the relationship between Microsoft Digital and the company’s employees.

“Now I can’t accidentally have guests in an internal-only group, which changes the dynamic. Employees can create their own Entra security groups now, without us having to worry that they’ll be inviting guests where they shouldn’t be.”

David Johnson, principal product manager architect, Microsoft Digital

For IT admins, the shift is from reactive remediation to proactive prevention. For employees, it means self-service action with security groups become viable again, without the security risks that made organizations reluctant to enable it before.

“Now I can’t accidentally have guests in an internal-only group, which changes the dynamic,” Johnson says. “Employees can create their own Entra security groups now, without us having to worry that they’ll be inviting guests where they shouldn’t be.”

Johnson underscores the broader ambition behind the shift, which is to allow employees to create and manage groups directly in Entra.

“A company that can unblock self-service action by its employees with confidence, knowing that there’s an additional level of protection—that’s very important,” he says.

Looking ahead: AI and the expanding policy surface

Labeling support for security groups is already being extended across the organization, with AI governance in mind.

Adding the ability to block agents from joining sensitive security groups is our next logical step. Guest membership is enforced via allow-to-add guest policy, but agents won’t join in the same way. Rather, we will set policies in Purview and then use labels to control if an agent can join a group.

The longer-term vision involves extending oversharing prevention beyond Entra itself. This will make it impossible (not just detectable) to accidentally assign a highly confidential resource to an unlabeled or inappropriately scoped security group. The foundation we’ve built with labeling in Entra is what makes this vital step possible.

“We want to get into the preventative aspect,” Johnson says. “The goal is to make it so it’s not possible to overshare in the first place.”

Key takeaways

Here are some tips as you consider ways to address how you manage your own security labeling practices:  

• Reuse existing labels—no extra setup required. Security groups reuse the same sensitivity labels already configured for Microsoft 365 Groups and SharePoint sites in Microsoft Purview, eliminating duplicate configuration and helping admins apply a consistent governance model across group types.
• Understand label immutability at launch. Unlike Microsoft 365 Groups, sensitivity labels on security groups are initially immutable—a deliberate design choice to ensure protections are enforced from the moment a group is created. Controlled label mutability will be introduced in a subsequent update.
• Know what’s in scope today. Labeling currently applies to static, non–mail-enabled security groups. Dynamic membership groups, mail-enabled security groups, and distribution lists aren’t supported at launch, so admins should plan accordingly.
• Shift from reactive cleanup to proactive protection. Label-driven membership controls prevent policy violations—such as unintended guest access—before they occur, reducing the need for post-creation audits and remediation.
• Enable safe self-service with guardrails. With labels enforcing access rules automatically, employees can create and manage security groups without increasing risk, restoring self-service without sacrificing control.
• Lay the foundation for future governance scenarios. Using sensitivity labels as the backbone of access policy creates a scalable framework that can extend to additional protections over time, including broader enforcement and emerging governance needs.

Try it out

Learn more about configuring security with Microsoft Entra.

Related links

• Learn how we’re using sensitivity labels to make Microsoft more secure.
• Get an overview of Microsoft Entra fundamentals documentation.
• Find out how self-service sensitivity labels to help our employees stay productive without exposing sensitive information.
• See what it took to transition our company to a Zero Trust security model.
• Explore how we’re boosting our Secure Future Initiative with a new approach to wired network security.

The post Transforming our approach to sensitivity labels at Microsoft with Microsoft Entra appeared first on Inside Track Blog.