Uguudei Bayaraa
AI Engineer & Full-Stack Developer
I build AI products powered by large language models, taking them from prototype to production. I keep things simple and dependable, so the software does its job and stays out of the way.
- Years building
- 5+
- Projects shipped
- 10+
- Focus area
- AI + Dev
A bit about who I am
A snapshot of my background, the things I care about, and the tools I reach for when building.
About Me
I'm a passionate AI engineer focused on LLM applications, retrieval systems, and full-stack product work. I love taking ideas from prototype to production: designing thoughtful UX, building robust data and inference pipelines, and obsessing over the small details that make software feel fast, reliable, and genuinely pleasant to use. That instinct for rigor comes from a strong mathematics background, built up by competing in math competitions from a young age.
Education
B.S. in Computer Science
The University of Chicago · Class of 2022
Sant High School
Class of 2017
Hobbies & Interests
Tech Stack
Where I've worked
A few of the teams I've helped build with, and the problems we solved together.
Software Engineer
E-Mongolia AcademyBuild and ship government mobile apps, including E-Mongolia, alongside Next.js portals and admin panels used at national scale. Own the release path end to end: CodePush over-the-air updates, App Store and Play Store submissions and compliance, and deploys to staging and production. Read the DevOps team's Docker configs and work within the CI/CD pipeline.
Software Engineer Intern
Steppe GroupShipped feature changes across a Flutter mobile app and its Appwrite backend, working both client and server side. Maintained and updated a Hugo static-site project.
Projects
A selection of things I've built, from AI systems to developer tools. Each links to its GitHub repo.
Atlas RAG
Production retrieval-augmented generation platform with hybrid search and evaluation dashboards.
PromptForge
Open-source toolkit for versioning, testing, and deploying LLM prompts at scale.
VisionFlow
Real-time computer-vision pipeline for edge devices, built with PyTorch and ONNX.
DevPilot
AI pair-programming assistant that reviews diffs and suggests refactors right in your IDE.
DataLoom
Streaming ETL framework that turns messy event data into clean analytics tables.
Inkwell
Markdown-native knowledge base with semantic search and AI-generated summaries.