Software Engineer — AI Agents
Projects
Non-technical staff ask questions in plain English. The agent finds the right columns across four internal systems, writes read-only SQL, and repairs only the clause that failed.
An ensemble classifier routes tickets across 18 intents. Three specialist agents answer in parallel, and low-confidence answers go to a human.
LLM résumé parsing with function calling and Zod validation, plus a Manifest V3 extension that fills applications on 300+ platforms.
Experience
San Jose, CA
Production agents for analytics and support: text-to-SQL over four internal systems and multi-agent ticket triage.
San Jose, CA
Job aggregation APIs, LLM résumé parsing, and a Chrome extension that autofills applications on 300+ platforms.
San Jose, CA
Full-stack social features in React, Next.js, and NestJS; cut feed queries from 40 to 3.
Henan, China
APIs, Celery pipelines, and dashboards for factory test data across 20+ device models.
M.S. in Computer Science
Santa Clara University · Santa Clara, CA
B.E. in Information and Communication Technology
Baekseok University · South Korea
Stack
About
I'm Haoyang — most people call me Lucas. I'm a software engineer who builds AI agents for production: the retrieval, tools, guardrails, and tests that make them dependable.
At Mil-Techs I build agents that let non-technical teams query internal systems and get support tickets answered in seconds. Before that, I built APIs, data pipelines, and dashboards at two startups and an electronics manufacturer.
M.S. in Computer Science, Santa Clara University