Data Scientist at HiLabs, building and shipping agentic systems, RAG pipelines, and production LLM applications. Outside of work, I enjoy composing music, exploring new restaurants, and catching a good sunset.
Lately, I've been spending a lot of time exploring domain-specific agent evaluation, post-training methods, and Knowledge Graph-powered AI systems. I'm particularly interested in building agents that can reason more reliably by combining LLMs with structured knowledge and domain context.
Most of my recent work has focused on healthcare data systems, where reliability, traceability, and domain knowledge are often as important as model quality.
arxiv-research-agent - An LLM agent that searches arXiv, reads papers, and produces cited research syntheses with rigorous evaluation.
| Project | What it does | Stack |
|---|---|---|
| FindMyGig | Agentic live music discovery - finds, ranks, and explains venues near you | LangGraph, Spotify API, Streamlit |
| Big Data Stock System | End-to-end S&P 500 anomaly detection pipeline | MongoDB, Airflow, Dask, Docker |
| Patient Survival Prediction | Classification model for ICU risk stratification | PyTorch, Scikit-Learn |
MS Information Management - University of Maryland BTech - IIT Bombay
Previously: Chicago Mercantile Exchange · Morgan Stanley
