Gagan Bansal
Principal Researcher at Microsoft
Greater Seattle Area United States
Summary
🤩
Rockstar🎓
Top SchoolGagan Bansal is a Principal Researcher at Microsoft Research AI Frontiers with a decade of experience bridging AI research and human-computer interaction, following a PhD from the University of Washington. He helped architect AutoGen, the open-source multi-agent framework that evolved into Microsoft’s official Agent Framework, and continues to co-lead influential agentic projects like Magentic-One and Magentic-UI. His hands-on contributions span core backend engineering—refactoring agent internals and improving assistant behavior—to practical tooling such as MarkitDown for document-to-Markdown conversion. Gagan combines rigorous academic training with product-minded engineering, routinely shipping tests and reliability improvements in widely used repos. Based in Greater Seattle, he’s known for translating complex multi-agent research into robust, developer-friendly systems. An understated strength is his focus on making agent behavior predictable and deletable—small design decisions that materially improve long-term maintainability.
10 years of coding experience
12 years of employment as a software developer
Doctor of Philosophy - PhD Artificial Intelligence, Doctor of Philosophy - PhD Artificial Intelligence at University of WashingtonIndian Institute of Technology Delhi (IIT Delhi)
Github Skills (16)
markdown10
openai-api10
markdown-it10
file-conversion10
backend10
zip-archive10
pytest10
7zip10
python10
zip10
testing9
document-conversion9
ppx8
autogen4
pdf4
Programming languages (5)
Github contributions (5)
microsoft/markitdown
Dec 2024 - Mar 2025
Python tool for converting files and office documents to Markdown.
Role in this project:
Back-end Developer
Back-end DeveloperContributions:80 reviews, 64 PRs, 54 pushes in 3 months
Contributions summary:Gagan primarily contributed to the `markitdown` project by implementing new features and fixing bugs related to file conversion. They introduced a `ZipConverter` to handle ZIP files, enabling the conversion of content within zip archives to markdown. Furthermore, they updated existing converters, such as the `PlainTextConverter` and `PptxConverter`, along with enhancements to existing unit tests. These changes reflect a focus on expanding the project's capabilities and improving its reliability.
autogenautogen-extensionlangchainmarkdownmicrosoft-office
microsoft/autogen
Sep 2023 - Mar 2025
A programming framework for agentic AI 🤖 PyPi: autogen-agentchat Discord: https://aka.ms/autogen-discord Office Hour: https://aka.ms/autogen-officehour
Role in this project:Backend Developer
Backend DeveloperContributions:485 reviews, 212 PRs, 362 pushes in 1 year 6 months
Contributions summary:Gagan's commits primarily focused on refactoring the `GPTAssistantAgent` within the `autogen` library, demonstrating a strong focus on backend development and agent behavior. They modified the constructor to handle instructions and the `overwrite_instructions` flag. They also improved the default behavior and added functionality to delete assistants, along with implementing tests to ensure the changes were functional and correct. These contributions reveal their expertise in the core logic of the agent's interaction.
agenticagentic-agiagentsaiautogen
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.