Hi, I'm Chonghan 👋
Lead Fullstack Developer at Apple, building enterprise workflow systems and agentic AI. I take LLM agents from prototype to production—multi-step LangGraph workflows, retrieval pipelines, and the latency budgets, alerting, and CI/CD that make them survive real traffic. I like turning ambiguous problems into systems a team can actually build on.
CW

About

Full-stack engineer and technical lead who ships multi-step LLM agents to production and keeps them running. I currently lead full-stack development of an internal enterprise workflow platform at Apple (Node.js, Vue 3), and built a 20-node LangGraph grading pipeline serving live traffic on GCE. My work spans agent orchestration, RAG and vector retrieval, backend APIs in Python and Node, and the reliability layer most AI demos skip—request budgets, failure-mode analysis, alerting on silent model failures, and health-gated deployment.

Work Experience

A

Apple

Jan 2026 - Present
Lead Fullstack Developer
Lead full-stack development of an internal enterprise workflow platform on Node.js services with a Vue 3 frontend, covering structured submissions, storage, editing, and process monitoring. Built responsive multi-field form flows that keep conditional fields, validation, and submission state consistent as a request moves through review. Maintain the admin configuration database and deliver Retool admin and scheduling pages so managers can make routine changes without engineering involvement. Partnered with the Tech Lead to design and roll out the engineering team's Claude Code agentic development workflow, and remediate security findings using internal tooling.
B

Better Buy
Part-time

Feb 2024 - Present
Software Engineer — Agentic AI Systems
Built and own a 20-node LangGraph agent that scrapes a product page, cleans it through four preprocessing stages, fans out to six parallel LLM analysis nodes, and aggregates them into a five-dimension product grade served to a Chrome extension. Designed conditional routing with a Perplexity web-search fallback for pages blocked by captchas or Cloudflare, and a two-stage DOM compression strategy that recovers product facts buried deep in retail markup. Bounded runaway latency with tiered per-call ceilings and a wall-clock budget per grade, made silent model-provider failures alertable after an outage that failed every grade for three days, and built the GitHub Actions pipeline that deploys both environments behind a health check.
H

HeartyByte Inc

May 2025 - Nov 2025
Software Engineer — Applied AI / LLM Platform
Designed and shipped an AI story-branching system on LangGraph that retrieves prior storylines, builds a summary knowledge base, generates candidate branch ideas, validates them for continuity, and expands the chosen branch into full narrative. Built retrieval pipelines over SERP API, OpenAI embeddings, and a vector store to extract trending keywords and feed structured data into automated SEO page generation. Applied LangSmith evaluation chains to score narrative coherence and character consistency, and rebuilt core UI/UX with Next.js tab-switching over a Firebase NoSQL backend. Real-time creator notifications raised user-to-creator interaction rates by an estimated 35%.
E

ENERlite Consulting, Inc

Feb 2022 - Feb 2024
Software Engineer
Developed a multi-process AWS pipeline for energy modeling and optimization, focusing on reproducibility, scalability, and deterministic results. Built data ingestion and validation systems that enforced parameter correctness and ensured consistent multi-step analysis workflows. Automated backend processing for simulations, created modular visualization tools, and containerized the entire pipeline with Docker to guarantee reproducible environments across AWS instances. Authored detailed documentation covering workflows, testing procedures, and expected outputs to support research teams and long-term maintenance.

Education

U

University of Southern California

2023 - 2025
Master of Science in Computer Science
U

University of California, Davis

2019 - 2023
Bachelor of Science in Computer Science and Engineering

Skills

LangGraph
LangChain
LangSmith
OpenAI API
RAG
Vector Stores
Agent Orchestration
LLM Evaluation
Python
TypeScript
JavaScript
C/C++
Bash
Go
Rust
Node.js
Vue 3
React
Next.js
Flask
Express.js
Docker
GCP
AWS
MySQL
PostgreSQL
MongoDB
Firebase
CI/CD
Linux/Unix
My Projects

Check out my latest work

I've worked on a variety of projects, from simple websites to complex web applications. Here are a few of my favorites.

Storio AI

An AI story-branching and SEO automation platform. A LangGraph agent retrieves prior storylines, generates and validates branch ideas, and expands them into full narrative, while retrieval pipelines over SERP API, embeddings, and a vector store drive trend-aligned page generation.

LangGraph
LangSmith
OpenAI
Vector Store
Next.js
React
TypeScript
Firebase
Python
Docker
Website

Better Buy

A Chrome extension that grades any product page across health, sustainability, ethics, durability, and pricing. A 20-node LangGraph agent scrapes the page, cleans it through staged preprocessing, and fans out to parallel LLM analysis nodes, with a web-search fallback for blocked pages and a wall-clock budget on every grade.

LangGraph
Python
Flask
Chrome Extension
React
MySQL
SQLAlchemy
GCP
GitHub Actions
Website

ENERtune

A multi-process AWS modeling pipeline with reproducible workflows, automated data validation, and Docker-based simulation environments to deliver deterministic energy-analysis results.

Python
AWS
Docker
PostgreSQL
React
Data Visualization
CI/CD
Website
Contact

Get in Touch

Want to chat? Feel free to reach out at chonghanvincentwang@gmail.com and I'll respond whenever I can.

LinkedIn