At Tap Talent.ai, I raised candidate-search precision from 31% to 95%+ by debugging ranking and matching across the UI, API, service, and database layers. I also hardened a multi-tenant form platform and delivered an identity verification flow using React, Node.js, MySQL, and AWS Rekognition.
At Handshake AI, I build cheat-resistant Docker coding benchmarks in Rust, Go, Python, TypeScript, and C, with isolated automated graders. I also turned correctness bugs in git, SQLite, Redis, and ninja into self-grading environments.
In my Digital Wallet System project, I designed a five-service fintech backend with a double-entry ledger and safeguards against double-spending. I also built an AI-based public health guidance system with a LangGraph agent, Qdrant RAG, and automated tests.
