I build the stuff
underneath the map.
10+ years as a geospatial software engineer — GIS platforms, spatial pipelines, and lately, AI agents that handle the boring parts. Born and raised in San Diego, still here.
I've lived in San Diego my whole life, and somehow ended up spending a decade building the invisible plumbing behind maps — GIS platforms, spatial pipelines, coordinate systems that all have to agree with each other before anything actually works.
I studied CS with a GIS focus at San Diego State, and I've been building enterprise mapping tools and data pipelines ever since — most recently, agents that handle the repetitive parts of engineering so I don't have to.
Outside of work: I climb, I mess around with digital art, I've built a couple of small video games nobody's played, and I read constantly. I also make weird ambient and house tracks by writing code instead of playing instruments — Strudel and TidalCycles, if that means anything to you. And I probably care about clothes more than a software engineer should.
In every one of these roles, I've owned architecture, mentored people, and worked directly with whoever actually needed the thing built — not just written code in a corner.
Route reads bottom to top, like a real topo — P1 is where I started.
Lead GIS Software Engineer
Leading architecture and development of a nationwide geospatial platform for permitting, site selection, and land development — including "Agent Flywheel," an internal agentic workflow system automating multi-step engineering tasks. Expanded jurisdiction coverage 67% while cutting processing time 55%.
Senior Backend Engineer
Built production geospatial REST APIs for routing and transportation analytics using FastAPI, PostGIS, and NetworkX. Reduced API response times from 20+ seconds to 3–5 seconds through caching and query optimization.
Senior Software Engineer — Geospatial
Shipped a computer-vision pipeline automating field survey classification at ~98% accuracy, replacing manual review. Architected cloud-native pipelines processing 60,000+ field records nightly.
Lead Geospatial Developer
Led enterprise GIS platforms supporting 200+ renewable energy sites across North America. Designed spatial data pipelines managing 10TB+ of operational and telemetry data; cut annual cloud costs by $250K.
Geospatial Developer
Built enterprise GIS applications supporting FEMA and Department of Defense emergency response programs using Django, PostGIS, and Leaflet. Automated spatial reporting, cutting generation time 75%.
Agent Flywheel
It started because I was tired of doing the same multi-step tasks by hand — data validation, pipeline scaffolding, writing the same reports over and over. I built an agent-based system that plans and executes those tasks itself, with a human checkpoint anywhere it touches production. It's been in daily use by the team since, and it's given people back hours a week they used to lose to busywork.
Zoning Data Scraper
Every city's GIS system disagrees with every other city's GIS system — different pagination, different coordinate systems, different schema quirks. I built auto-detected pagination, coordinate reprojection, and WKID alias resolution into one reusable ingestion package, backed by DuckDB storage with upsert logic to keep it current, plus a 40-test suite so it doesn't quietly break. It covers Austin, Tempe, and Portland today. It's proprietary — built for internal/client use, so no public repo, sorry.
Field Survey CV Pipeline
Field survey review used to be a fully manual bottleneck — someone had to eyeball every record before engineering could move forward. I trained and deployed a CV classification model straight into the existing FastAPI/PostGIS pipeline, so records get classified automatically as they come in. It's been running at ~98% accuracy in production, and it took a real chunk of manual work off the team's plate.