I'm a full-stack AI engineer in Sunnyvale, California. I build production AI products end to end: the data pipelines underneath, the LLM and RAG systems in the middle, and the web apps people actually use.
At Uber I built Spark and Kafka pipelines processing 200M+ records a day and cut $290K a year in compute. I've also evaluated frontier-model output for a top AI lab, shipped a multimodal RAG system that's live in two products, and founded Workify. UC Berkeley CS, full-ride Regents' Scholarship.
Previously at
Mercor
Intro
By the numbers · at Uber
Experience
Founder & Engineer
Founded a fleet-maintenance platform with an AI manual assistant. Own the product, architecture and engineering end to end.
Software Engineer I, Growth & Marketing Data
Owned end-to-end Spark and Kafka projects for growth and marketing analytics on 2B+ users. Built PySpark ETL pipelines and Kafka topics handling 200M+ real-time location rows a day, cutting personalization latency from 1 day to under 6 hours. Data-validation work saved $290K a year in compute. Recognized by a Staff Engineer for ownership and data-quality initiatives.
Software Engineering Domain Expert (Contract)
Provided software engineering expertise to improve frontier models for a top AI lab, evaluating and refining complex technical outputs.
Software Engineering Intern, Anomaly Detection
Parallelized Cadence workflows in Go across 370 metric groups, cutting anomaly detection from 1 hour to under 15 minutes. Moved a gRPC API layer from Python to Go.
Data Engineering Intern, Business Intelligence
Automated Salesforce-to-Snowflake ETL and Power BI dataflows, removing 4 hours of manual reporting a week. Built scrapers covering 560+ competitor models.
Undergraduate Research Apprentice
Built Python pipelines to clean, segment and analyze 150,000+ unstructured JSON files for network research. Co-built a Swift iOS app (4,500+ lines) used by 1,000+ people.
Founder · WorkifyWorkify gives maintenance teams an AI repair assistant that answers from their equipment's service manuals, text and diagrams, plus team accounts, seats and billing plans that grow from one technician to the whole fleet.
Inside Workify
My role Full Stack Developer: Web App, Billing, and AI Chat
Workify helps companies run fleet maintenance in one place. The web app has company and team accounts with seats, Individual, Team, and Professional plans plus a custom Enterprise plan, and Stripe billing. It also includes an AI chat that answers repair questions from equipment service manuals, powered by my RAG backend.
My role AI Engineer: built the RAG pipeline and FastAPI backend
An AI assistant that answers questions about machine service manuals using both the text and the diagrams in the PDFs. Users can ask in English, Burmese, or Japanese and get answers in the same language. It runs as an always-on FastAPI server that keeps everything in memory for fast responses, and it powers the manual chat in two live products: the Seinn Yaung So marketplace and the Workify app.
My role Full Stack Developer: marketplace site and AI integration
A marketplace site for Seinn Yaung So, a farm machinery dealer in Myanmar. Customers browse and list spare parts, view products, and ask questions about machine manuals. I connected the site to my RAG backend, so the AI answers from the real service manuals, diagrams included, in the customer's own language.
Other work
My role Full Stack Developer
A web app that shows employees their stock options, RSUs, and bonus payouts as they vest. It tracks every grant and vesting schedule, shows live pricing, and keeps a tamper-proof history of every change. It shares a database with the main Workify app.
My role Full Stack Developer
An internal toolkit that automates the team's admin work. One tool converts spreadsheet cost reports into the accounting system's import format. Another researches equipment dealer contacts and drafts outreach emails. The same tools run as a desktop app, a web dashboard, or from the command line.
My role Full Stack AI Developer
Our team won 2nd place and a $3,000 prize at the Masumi Hackathon in San Francisco. Aeruku is an autonomous AI sales agent that handles outbound sales with minimal human input. It finds leads through Apollo, LinkedIn, and web scraping, writes personalized messages using web context from the Brave API, and adjusts outreach based on engagement.
Skills
B.A. Computer Science, 2022 – 2025 · Regents' Scholarship (full ride)
English (native) · Burmese (native) · Chinese (professional)
Contact
Want a Workify demo for your fleet, need an engineer for an AI or data project, or hiring? Send me a note and I'll get back to you within a day.