Rethinking device management internally at Microsoft with AI

At Microsoft, we’re leaning on AI to enhance our internal device management strategy.

AI is helping us to simplify and improve the experience our employees have with their devices by predicting and auto-remediating issues, supporting proactive solutions, and enhancing the operating system’s look and feel.

As hybrid work becomes the norm—and the expectation—for our employees, how we in Microsoft Digital, the company’s IT organization, give them access to the tools they need to successfully innovate, create, and collaborate has evolved. Employees want a dynamic, device-agnostic experience that focuses on providing them with the data and tools that they need from almost any location, using a wide variety of devices, including PCs, laptops, tablets, and smartphones.

“We’re investing in AI-powered predictive maintenance and intelligent troubleshooting to reduce friction in device management,” says Daniel Manalo, a principal service engineer at Microsoft Digital. “We’re using AI and machine learning to help us schedule essential maintenance tasks and fix errors and performance issues autonomously.” “This is reducing downtime, prolonging device lifespans, and ensuring our employees have a consistent and productive experience by avoiding problems and errors.”

Manalo and his team are investigating ways to use AI to analyze device settings, network activity, vulnerabilities, and user behavior, enhanced with demographic data and location metadata to offer relevant solutions for common and emerging device problems.

While device management focuses on the employee experience, Manalo reminds us that Microsoft Digital support teams can benefit greatly as well.

“We want to help our support team be more productive through quicker decisions about device replacement, software updates, capacity increases, and other common support scenarios,” Manalo says.

Using AI to reduce friction in device management

Our employees use a wide variety of devices as their primary productivity tools to access their work and succeed in their roles. Our responsibility at Microsoft Digital is to ensure that each of our employees can be productive and connected to Microsoft tools and corporate data, regardless of the device they use.

There are more than 750,000 devices in use at Microsoft including Windows, Android, iOS, macOS, and Linux devices. Approximately 60% of these are Windows devices, while iOS, Android, and macOS account for the rest. Of these devices, approximately 45% are personally owned employee devices, including phones and tablets. Microsoft empowers employees to use the managed devices that enable them to be their most productive.

Microsoft Intune supports the modern management model at Microsoft. Intune provides cloud-based device management capabilities across Windows, Android, iOS, macOS, and Linux devices. Devices are registered in and authenticated by Microsoft Entra ID. Because it’s cloud-based, Intune removes the dependency on the local network, and managed devices can connect across the internet from anywhere. Modern management includes and supports both corporate and personally owned devices, including mobile devices.

However, even with the benefits of modern management, we recognize that there’s room for improvement.

Employee productivity and sentiment are directly affected when the condition of their device and the underlying infrastructure deteriorates. Unexpected reboots, application crashes, emerging vulnerabilities, and compatibility problems all negatively affect the employee experience. The situation is further aggravated by potentially long wait times with Helpdesk to resolve support tickets. And not all issues are reported by employees.

Our support teams spend large amounts of time manually pulling together data and insights to make long-term preventative decisions about device replacement, software upgrades, capacity increases, and more. They don’t have aggregated views with device health insights and the toolsets to analyze patterns and trends to reduce their decision-making time or increase their confidence that they made the right choices.

Put simply, we understand that our modern management processes have gaps, and we’re filling those gaps with AI-powered tools and services.

“The goal is to make the device smarter,” says Senthil Selvaraj, a principal product manager at Microsoft Digital. “We want the device—and the services that support it—to be intelligent and able to predict or detect issues on the device and self-remediate.”

Selvaraj’s team is focusing on using AI to provide detection and remediation in a way that prioritizes and respects the employee experience. “We don’t want our tools to consume a lot of local resources when the employee might need those resources for other tasks,” he says. 

Selvaraj says the focus is on creating a productive and frictionless device experience at all times. “We don’t want additional load on the device, so we want to make sure we’re running automated remediations at the right time without any impact to users,” he says.

Integrating AI for proactive maintenance and issue resolution

Disruptions in our enterprise device and infrastructure environment increase support costs and reduce employee productivity. When employees encounter an issue, they must stop whatever they’re doing and either fix the issue or report it to the IT helpdesk. Long resolution times for support tickets and lack of detailed insights for IT administrators further impact employee productivity and increase our support costs.

As Customer Zero for Microsoft, we’re developing and implementing AI-powered solutions that will simplify and improve the employee device experience.

“We’re developing an AI and automation solution that monitors, predicts, and resolves device and infrastructure issues for employees and IT admins,” says Dave Rodriguez, a principal product manager on the Frictionless Devices team in Microsoft Digital. “The solution uses data from our enterprise devices, such as laptops, network devices, sensors, and meeting room equipment to find and fix problems before they impact the users.”

The team is building capabilities that will actively solve IT challenges and lighten our employees’ cognitive load by proactively delivering solution-focused notifications and recommendations, while also addressing their queries about their device experiences.

“Using generative AI and natural language understanding, we’re providing IT administrators with a conversational AI experience,” says Pandurang Kamath Savagur, a senior program manager with Microsoft Digital. “This is enabling them to query patterns, observe analytics, and get recommendations across the device and infrastructure environment to manage and prevent disruptions.”

Specifically, our new solution provides an automated, AI-driven device experience by:

• Mining the vast telemetry that we capture across our devices and infrastructure to ground AI-based remediation and automation.
• Aggregating and collating the anomalies detected across devices and infrastructure to identify root causes of issues and impacted areas.
• Combining near real-time telemetry and historical anomalies and issues to predict and fix issues across the enterprise device landscape before they start negatively impacting employee productivity.
• Providing IT administrators with deep insights into the health and performance of enterprise devices by analyzing signals and demographic data to detect anomalies and proactively identify issues.
• Integrating with existing Microsoft products such as Microsoft Intune and Microsoft Teams that support and supply the employee device experience.

We’re excited to realize the potential of this solution for device support as we extend implementation and roll out to the larger device community at Microsoft.

Improving device security results with Microsoft Copilot for Security

Our security and IT teams are using Copilot for Security to protect at the speed and scale of AI, while remaining compliant to responsible AI principles.

Copilot for Security integrates directly with Microsoft Defender XDR, Microsoft Sentinel, Microsoft Intune, and many other security-relevant data sources to create a unified experience for large language model (LLM)-powered prompting, grounded in the data from integrated solutions.

We’re moving toward a holistic approach in which we enhance common use cases with Copilot for Security capabilities, including:

• Incident summarization. Copilot for Security is helping us gain context for incidents and improve communication across our organization by using generative AI to swiftly distill complex security alerts into concise, actionable summaries to enable quicker response times and streamlined decision-making.
• Impact analysis. Copilot for Security uses AI-driven analytics to assess the potential impact of security incidents, offering insights into affected systems and data to prioritize response efforts effectively.
• Reverse engineering of scripts. It helps us simplify malware investigation with automated reverse-engineering for scripts so every analyst can understand the actions executed by attackers. It also analyzes complex command-line scripts and translates them into natural language with clear explanations of actions.
• Guided response. We receive actionable step-by-step guidance for incident response, including directions for triage, investigation, containment, and remediation. Our support team also receives relevant deep links to recommended actions that allow for quicker response.

Copilot in Windows and Microsoft 365 Copilot

Windows PCs are still the primary working environment for our employees. and are critical to our business. Copilot in Windows is an AI-powered assistant that’s built into the Windows operating system and uses advanced machine learning and natural language processing to provide intelligent suggestions, automate tasks, and integrate seamlessly with Microsoft services. 

Copilot in Windows brings Copilot to the taskbar, providing a natural language companion ready to assist our employees. It’s transforming how Microsoft employees work, allowing them to focus on strategic and creative tasks. 

Integration with Microsoft 365 Copilot is at the core of our Copilot in Windows deployment at Microsoft Digital. 

“Microsoft 365 Copilot provides the most productive integration for our employees,” says Harshitha Digumarthi, a senior product manager at Microsoft Digital. 

Integrating Microsoft 365 Copilot with Copilot in Windows ensures adherence to enterprise security, governance, and trust standards. It also gives Microsoft employees a generative AI tool grounded in our enterprise data to get relevant, authoritative, and helpful answers and content directly from the Copilot in Windows interface.

Digumarthi’s team is also dedicated to understanding how updates influence user productivity and experience.

“We’re asking the important questions,” Digumarthi says. “How can Copilot in Windows and Microsoft 365 Copilot enhance productivity? Will it introduce any changes that might lead to confusion?” 

Looking forward

We’re constantly examining new ways to use AI to extend our device management capabilities and are working toward integrating Copilot for Security more deeply into our device management and security practices. We’re also deploying our automated AI device management solution to a wider set of devices as we continue to refine existing features and develop new ones.

AI is making us rethink device management at Microsoft Digital. We’re using AI to enhance the user experience, predict and resolve issues, support proactive solutions, and improve security outcomes. This integration spans the entire device spectrum, from the employee experience to the services and tools that facilitate device management. We’re only just starting to uncover the possibilities in AI to simplify and improve device management and empower our employees to work from anywhere, on any device. We look forward to growing our device management capabilities alongside AI advancements in the future.