Ankur Makwana

About Ankur Makwana

I build large-scale generative AI agent platforms and help define how humans and AI collaborate inside the enterprise.As a Principal Product Manager on Microsoft Copilot Studio, I lead cross-functional teams of AI researchers, engineers, and designers to ship foundational capabilities that power the world’s most widely used enterprise agent ecosystem. My work focuses on making agent creation intuitive, reliable, and grounded, even in the messy reality of enterprise data.I’ve driven major innovations in low-code agent building, including the development of a meta-agent that enables information workers to create production-ready agents using natural language. I’ve also led end-to-end work on retrieval quality, grounding, data connectivity, and evaluation frameworks for agents that integrate with 3rd-party tools and knowledge sources outside the Microsoft ecosystem.My product philosophy is simple: turn cutting-edge research into dependable, scalable systems that ship. I’m known for translating ambiguous problems into strategy, aligning diverse teams, and executing with intensity and care while delivering incremental customer value.Right now, I’m focused on the next frontier of improving the response quality, articulation, trustworthiness, and citation rigor of enterprise AI agents that are connected to Microsoft as well as 3P data sources by rethinking data access, indexing, retrieval, and evaluation from the ground up.Let’s connect if you’re building agent platforms, designing grounding and retrieval systems, or exploring new ways to evaluate and improve AI-powered reasoning. I’m always excited to learn from others and help push the boundaries of human-agent collaboration.

Experience

1. Principal Product Manager Copilot Studio Ai Agents

   Microsoft

   Aug 2025 — Present · Seattle, WA, US

   Improving enterprise agent response quality across top external data sources by redefining how Copilot Studio agents handle data retrieval, grounding, reasoning and citation rigor for non-Microsoft ecosystems.Designed the end-to-end evaluation strategy by identifying what customers value (relevance, correctness, articulation, trust, citations), building ground-truth datasets representing real-world scenarios, and creating LLM + human eval frameworks to measure and quantify response quality.Drove systematic quality improvements through targeted experiments on retrieval, grounding, and articulation improving the baseline agent accuracy and trustworthiness for the Top used external data sources.

Education

• National Institute of Technology Surat

  Bachelor of Technology (B.Tech.), Electronics Engineering

  2006 — 2010

Skills

• Predictive Analytics
• Mobile Search Advertising
• Online Advertising
• Big Data
• Analytics
• Linear Regression
• Sas Programming
• Customer Analysis
• Data Analytics
• Team Management
• Management Consulting
• Sql
• Microsoft Excel
• Segmentation
• Marketing Analytics
• Business Process Improvement
• Descriptive Analysis
• R
• Web Analytics
• Regression Analysis
• Business Strategy
• Quantitative Analytics
• Microsoft Sql Server
• Business Insights
• Cluster Analysis
• Offshoring
• Global Delivery