Selected Work
CTD RAG Chatbot
Retrieval-augmented generation pipeline for clinical trial documentation. Improved retrieval precision by 60% and increased observability by 40% through LLM-as-a-Judge evaluation infrastructure.
- Next.js
- TypeScript
- MongoDB
- Langfuse
- LangChain
- OpenAI
2nd Place, MetLife Hackathon 2025
Chef Jul
AI-powered meal planner with personalized nutrition intelligence built in 48 hours. Parallelized LLM pipeline cut processing time by 40%; multi-agent workflow enables ~2s preference adjustments.
- Firebase
- React
- Node.js
- LLM Orchestration
About
I focus on writing performant, maintainable, production-ready code and shipping systems that hold up under real-world use. At Code the Dream I work on an AI-powered benefits chatbot built on Retrieval-Augmented Generation, where I own evaluation infrastructure end to end — from LLM-as-a-Judge scoring pipelines and Langfuse observability to the streaming chat interfaces users actually see.
ResumeEducation
React, Node.js, Python and Advanced Practicum
Code the Dream · 2024
Software Engineer Immersive Program
Hack Reactor · 2018
Location
San Francisco, CA