Building Portfoliq · SWE @ AWS

Luke Chen

I build and ship production systems across hardware, ML, and full-stack SaaS. From concurrency primitives and embedded firmware to multi-agent AI systems and shipped SaaS.

See selected workGitHubAtlanta, Georgia
Scroll

Selected Work

Things I've shipped

Experience

The build log

Sep 2024 — Present

Software Engineer

·Amazon Web ServicesNow
  • Designed and shipped a production multi-agent RAG system — Claude-powered agents over hybrid semantic and vector retrieval on a serverless, event-driven pipeline.
  • Led an AWS Marketplace procurement migration, moving 10,000+ ML models and services into a unified system supporting 5+ pricing models with zero critical post-launch issues.
Mar 2024 — Sep 2024

Software Engineer

·Visa
  • Led a full-stack migration from .NET Framework to .NET 8 across a web application used by card issuers.
  • Pioneered generative AI at Visa — adopting GitHub Copilot and OpenRewrite to automate code refactoring on a stack with no prior automation tooling.
May 2023 — Aug 2023

Software Engineer Intern

·Amazon
  • Built a Java verification tool evaluating UI rendering accuracy across Amazon's homepage and 600M+ product pages, plugged into EC2, Lambda, and CloudWatch for real-time monitoring.
May 2022 — Aug 2022

Software Engineer Intern

·Amazon
  • Shipped a full-stack management portal for Amazon's 9.7M retail sellers — DynamoDB schema, Java backend, and a JS/JSP frontend with REST + AJAX — end-to-end in 12 weeks.
Jan 2022 — May 2022

Computer Hardware Teaching Assistant

·Georgia Tech
  • Taught breadboard wiring, FPGA programming in VHDL, and oscilloscope-based signal analysis to 129 students.
Jun 2021 — Aug 2021

Software Engineer Intern

·UKG
  • Built a back-end tax-return processing tool in C# using UiPath's RE Framework with AES encryption — an estimated $303,866 annual ROI.

About

Builder, metal to model.

Software engineer at AWS and ML master's student at Georgia Tech. I work from the metal up — concurrency primitives and embedded firmware on one end, multi-agent AI systems and shipped SaaS on the other. The throughline is shipping things real people depend on.

I've also packaged a tangle of CRM, ops, and automation tools into a turnkey “business in a box” for first-time operators — onboarding four partner-run locations with zero added headcount and roughly $1,000/month back in each operator's pocket through negotiated vendor pricing.

Education

M.S. Computer Science — Machine Learning

Georgia Institute of Technology

Jan 2024 — Dec 2026

B.S. Computer Engineering

Georgia Institute of Technology

Aug 2019 — Dec 2023

3.90 GPA · Dean's List · Nordson BUILDS Scholar

Currently

Engineering at AWS by day, finishing an ML master's, and growing Portfoliq with real customers.