Microsoft · 2024
Operations Agent
LLM-powered agentic system for operational intelligence. Built end-to-end: reward signals, evaluation framework with auto-raters, and production deployment on Fabric.
Applied AI Engineer · Ph.D. Mathematics
I build AI systems
that ship.
From reward models to agentic systems, I bring mathematical rigor and genuine curiosity to the craft of making AI work in production. Currently leading AI product development at Microsoft Fabric.
Selected Work
View all → Microsoft · 2023–24
Anomaly Detector
AI-powered anomaly detection for real-time streaming data. Scaled to handle enterprise-grade telemetry with low-latency inference and interpretable outputs.
Microsoft · 2024
Operational Discovery
Automated discovery of operational patterns across time-series data. Combines statistical methods with LLM-generated insights for actionable intelligence.
The arc
Math → Data Science → Applied AI
I started as a mathematician, drawn to the elegance of proofs and abstraction. That foundation taught me to think precisely about uncertainty — which turned out to be the perfect training ground for building AI systems at scale. Across five major tech companies, I've evolved from modeling to engineering, always asking: how do we make this actually work?
More about me →