AI Interaction and Learning

The AI Interaction and Learning (AIIL) group in Microsoft Research focuses on building and analyzing symbiotic AI systems, where knowledge workers and AI agents collaborate to produce outputs of ever-increasing value. To achieve this, our current projects topics include:

Semantic Telemetry and User Modeling: Assessing how people use copilot systems at scale and measuring value to the user. This includes work on measurement, evaluation, and synthetic data generation.

Interactive ML and Continual Learning: Utilizing interaction traces to extract signals (eg. preferences, feedback) and improve the performance of the AI systems. This includes work on personalization, recommendations, prompt optimization, fine-tuning, and alignment.

The group brings together expertise in Microsoft Research across different academic subdisciplines to work together on this mission including artificial intelligence, machine learning, natural language processing, recommender systems, information retrieval, and computational social science. Together with our product partners, our team has developed technology that has improved product offerings in Copilot, Office, Bing, and XBox.

We currently have applications open for interns and full-time researchers on our career opportunity page. We are always interested in working with individuals who have a strong interest in pushing the boundaries of learning and reasoning with foundation models and who are ready to think broadly, focus deeply, and work as a team to advance the state-of-the-art. We invite you to learn more about us on our respective profile pages.