University of Southern California
Viterbi School of Engineering · Los Angeles
M.S. in Computer Science
EXPECTED MAY 2027Software engineer · Open to SDE opportunities
👋 HELLO, I’M
Jason Bu.
A software engineer.
A curious builder.
I’m Jason Bu, a computer science master’s student at the University of Southern California and a computer science graduate of the University of Utah. I focus on backend development, AI agents, and infrastructure, and enjoy turning complex problems into practical solutions.
During software engineering internships at MW Polar and Sunline Tech, I worked on AI workflow automation, banking-system performance, and CI/CD pipelines. Through Maestro, Revieu, and Spreadsheet Engine, I also explore multi-model orchestration, backend architecture, and software systems design.
Outside of coding, I enjoy traveling, cooking, researching AI, and analyzing the stock market. I’m seeking SDE opportunities where I can work with a team to build useful, reliable products.
I enjoy breaking down complex problems and using code to solve real needs.
1.35B
My Codex tokens used
LOCAL SNAPSHOT / 2026-10-01About this number
1,352,732,249 tokens recorded in retained local Codex sessions from March 4 to October 1, 2026. Includes cached input and subagent usage; excludes inherited fork counts. A dated snapshot, not a live counter.
01 / Education
My education.
Computer science, from Salt Lake City to Los Angeles.
University of Utah
Price College of Engineering · Salt Lake City
B.S. in Computer Science
MAY 2025Meet Jason
A little more about me.
My background, the work I care about, and what I’m looking to build next. A short introduction is on the way.
INTRODUCTION / COMING SOON
02 / Experience
My journey so far.
Turning operational bottlenecks into faster workflows, resilient services, and releases teams can trust.
30 secFastest PO processing
100K+Daily users supported
134Automated regression tests
35%Lower inquiry latency
MAY — JUL 2026
Software Engineer Intern
MW Polar · Los Angeles, CA
- Built an AI document-extraction pipeline for purchase orders, inventory checks, duplicate detection, and validated record storage; cut processing time from 5–20 minutes to 30 seconds–3 minutes across 30 test cases.
- Led an LLM sales-operations agent from warehouse coordination to shipment tracking, removing three manual coordination steps per request.
- Built CI/CD pipelines with GitHub Actions, Jenkins, Docker, and AWS EC2; automated 134 regression tests and standardized releases.
JUN — AUG 2025
Software Engineer Intern
Sunline Tech · Shenzhen, China
- Designed a Redis cache-aside layer with TTL controls for loan inquiries, lowering latency by 35% in a service handling 100K+ daily users.
- Built Java and Spring Boot quick actions for a banking app's customer-service screen, reducing clicks to common account details by about 45%.
03 / Selected work
Selected projects.
Backend engineering, AI orchestration, and product thinking — grounded in real constraints.
Multi-model orchestration · In development
01Maestro
A Python orchestration core for model routing, event-based recovery, budget accounting, and controlled agent execution.
Engineering details
Implemented configuration, event storage, artifact storage, budget accounting, and policy modules. The complete execution loop is still in development; performance targets await benchmark evidence.
SQLite WALEvent store
PolicyPermission boundaries
ReplaySnapshot recovery
BudgetReserve & settle
Campus discovery · Team project
02Revieu
An AI-enhanced discovery and e-coupon platform. Led backend technical design within a six-person team.
Engineering details
Built pgvector retrieval for merchant-grounded review drafts and a Redis/PostgreSQL coupon workflow. Resume-reported internal testing: draft acceptance rose from 61% to 83%, with 5K+ concurrent claim requests and 20% fewer duplicate claims.
61 → 83%Internal draft acceptance
5K+Concurrent claims
−20%Duplicate claims
GoBackend service
Software systems · Spring 2024
03Spreadsheet Engine
An Excel-style C# spreadsheet model with formula evaluation and dependency-graph recalculation.
Engineering details
Implemented cell validation, XML save/load, and a graphical front end. Added tests across formula, dependency, and spreadsheet layers.
C# / .NETCore engine
GraphDependency updates
XMLSave & load
TestsLayered validation
04 / Toolkit
Technical expertise.
05 / Off screen
Curious by nature.
New places, good food, market patterns, and the next AI paper worth reading.
01Travel & new places
02Cooking
03AI research
04Stock market analysis
06 / Say hello
Let's build something reliable.
Open to software engineering opportunities and thoughtful conversations.