At Expedia Group, I build AI-native discovery and recommendation capabilities for travel products. I architected an agentic AI evaluation platform using Python, LangGraph, tool calling, guardrails, and LLM-as-judge workflows to assess groundedness, hallucination, and task-level behavior.
I also built RAG pipelines that combine embeddings, hybrid search, vector retrieval, and re-ranking for context-aware discovery and recommendation. For travel discovery, I designed multi-agent workflows using LangChain, structured outputs, function calling, and context engineering.
At Microsoft, I developed production REST APIs and distributed services in Python, integrating SQL-backed data flows, asynchronous processing, and ML components. I also built ML-enabled capabilities and experimentation pipelines involving PyTorch, TensorFlow, and A/B testing.
Earlier, at Lytmus Inc. and Amazon, I developed backend services, APIs, data-processing components, and experimentation pipelines. Across these roles, I’ve worked on automated testing, CI/CD, deployment safeguards, and production troubleshooting.
