Master's Graduate in Health Data Science at Harvard T.H. Chan School of Public Health
Global mindset shaped by experiences across Harvard, Dartmouth, and Hawaii

I’m a data scientist and AI/LLM enthusiast passionate about building solutions at the intersection of public health, machine learning, and generative models. I’m particularly excited about using language models (LLMs) and retrieval-augmented generation (RAG) to assist real-world decision-making in healthcare and other high-impact domains.

These are a list of ideas for projects once I complete the current projects. The goal is to build the habit of creating production-style code with logs, tests, containerization, etc. Each of the projects will likely have a learning log that summarizes lessons in coding, concepts, and the majority of mistakes learned along the way in the building process.

LLM-RAG: Self-consistency, Multi-hop, Re-ranking, Structured Context, Fusion techniques, Query Rewriting, Hierarchical.

LLM-Tuning: Fine, adaptor, prefix, prompt, instruction

LLM-Evals: Multi-LLMs as a judge and use concensus to proxy human, Ground truth labels vs Judge, Adversial stress testing (deliberatlty plant wrong/ not factual answers to see if LLM gets it correc)

Multimodal: text and photos to some degree

Optimizaitons: caching, traffic, and performance

Misc: MCP, SGLang, vllms, distillation, pruning, batching