Greater Seattle Area
I lead an applied science and data science team focused on building the Copilot Offline Evaluation Platform, enabling rigorous, scalable evaluation of AI-powered experiences across Microsoft 365. My role drives innovation in model assessment, synthetic data generation, and metrics frameworks to ensure quality, reliability, and user trust in Copilot solutions
● Sep'19 - Feb'21: M365 Core ● Aug'17 - Aug'19: build machine learning models and conduct causal inference analysis to support marketing decisions and initiatives for Game Pass and Xbox Console. ● Jun'16 - Jul'17: compute and storage demand forecast for Microsoft Cloud Infrastructure and Operations.
Built machine learning models for short term power price prediction in the energy industry (locational marginal pricing); developed nonlinear optimization model for optimal day-ahead power trading.
Worked cross functionally with supply chain and finance leaders on optimal crop production planning and resource allocation for long horizon R&D financial budgeting (through simulation & optimization).
Conduct research in the field of computational optimization, particularly in the area of combinatorial and global optimization, and non-convex quadratic programming.