About
Gagan Bansal (he/him) is researcher at Microsoft Research AI Frontiers where he conducts interdisciplinary research on Artificial Intelligence and Human-Computer Interaction.
Prior to joining MSR in 2022, he graduated with a Ph.D in Computer Science from University of Washington, Seattle where he was advised by Dan Weld (opens in new tab) and worked as intern at Microsoft Research for multiple summer under the guidance of Besmira Nushi, Ece Kamar, and Eric Horvitz.
Since 2023 he was one of the technical leads of AutoGen (opens in new tab), a widely used open-source framework for developing multi-agent applications, before it transitioned to product in mid 2025 and now forms the backbone (opens in new tab) of Microsoft’s official Agent Framework (opens in new tab).
Since then, he has co-led several other widely popular open-source agentic project including Magentic-One a state-of-the-art multi-agent team for solving complex task, its GUI friendly version Magentic-UI, MarkitDown (opens in new tab) a tool for converting documents to markdown, and Magentic-Marketplace (opens in new tab), an environment for understanding societies of agents in two-sided agentic markets.
He has led Microsoft’s early thinking on improving human-agent interaction, checkout Challenges for human-agent communication.
Featured content
Introducing AutoGen v0.4
This video provides an in-depth look at AutoGen 0.4, a significant upgrade to Microsoft's open-source framework for multi-agent AI. Gagan Bansal, explains how this version addresses user feedback by introducing a new layered architecture designed for greater modularity, flexibility, and scalability
Why Human-Centered Agents are Hard to Build?
This talk by Gagan Bansal from Microsoft Research AI Frontiers focuses on the challenges of building human-centered AI agents.
GitHub Star History
At Microsoft, I've co-led several influential open-source projects. Checkout their start history!