Applied AI Engineer

Hi, I'm Geon.
I build AI systems that make sense of high-stakes data.

From viral genomes to live emergency operations — same instinct, different domains: find the signal, structure it, make it usable under pressure.

View my work ↓

About

I build systems that turn messy, fast-moving data into decisions worth trusting. That's meant processing billions of sequencing reads to trace how a virus mutates, and it's meant reconstructing an emergency incident in real time so a commander can see what's actually happening.

Currently working as an Applied AI Engineer, following an MS from UC Berkeley's School of Information and research in viral genomics at the UW Virology Lab.

PYTHON PYTORCH LLMS NLP AWS R NEO4J GENOMICS SQL

Experience

2026 — Present

Applied AI Engineer

Rate (formerly Guaranteed Rate)

Training and deploying AI-assisted applications and pipelines for mortgage operations. Currently building document extraction systems for purchase contracts, helping underwriters and mortgage experts get accurate data faster.

2023 — 2026

Bioinformatician

UW Virology Lab

Processed 30B+ sequencing reads for viral genomics and phylogenetics, built ML models for biomedical imaging, shipped R Shiny dashboards on AWS, and co-authored peer-reviewed publications.

2024 — 2026

M.S., Information Management & Systems

UC Berkeley, School of Information

Thesis became the seed for Inference, an AI decision-intelligence platform for emergency command — see Previous Projects.

Live Projects

ACTIVELY MAINTAINED

WatchBot

Self-hosted restock alerts, built end to end

A restock-alert tool for friends chasing limited retail drops. Multi-account picker, price-threshold alerts that flag scalped listings above MSRP, and push notifications the moment something's back in stock. FastAPI backend, background polling, htmx frontend, containerized and self-hosted.

See previous projects →

Contact

Have a question or want to work together? Reach out directly.