experience

helixdb (yc x25)

• developed a next.js dashboard allowing users to query databases, visualize data, and browse database schemas
• implemented a Gemini agentic retrieval system with MCP tools in Python, featuring supervisor-sub agent architecture for graph traversal, vector search, and property filtering on 1.5+ million nodes and vectors
• implemented an agent using Claude Agent SDK in TypeScript
• programmed an 8-phase ingestion pipeline inspired by OpenAI's cookbook that transforms raw data into a temporal knowledge graph through parallel extraction, embeddings, entity resolution, predicate deduplication, and invalidation checking
• created an interactive CLI application in Rust that achieved 2,800+ all-time downloads
• built REST endpoints into the Rust database to allow visualization of data
• developed a knowledge graph planner and retriever agent that performs temporal queries
• deployed a full-stack app for the agents with Next.js frontend and Hono.js backend on AWS Lambda

uc san diego health

• developed and maintained R scripts for data cleaning, manipulation, and transformation, enabling efficient preparation of datasets for research and analysis
• created reusable R templates and functions for data analysis and reducing errors in research outputs
• conducted statistical analyses on large health datasets using Stata

civilpark international

• built a web app using Angular, C# (.NET Framework), and MySQL
• added 18 different modules for complex wind calculations (non-uniform high-rises, factories, varying floor heights, different terrains)