Work With Me
I'm an independent senior engineer for maps, geospatial, and data systems. Small and mid-sized businesses hire me to build the map or data feature their product depends on, fix the one that stopped scaling, or make a slow Postgres database fast — backend, data, infrastructure, and the map layer, owned by one engineer.
One engineer across the whole stack: fewer handoffs, fewer meetings, one person who takes a problem from framing to production.
What I Do
- Maps and geospatial systems — Spatial APIs, vector tiles, geocoding and place search, PostGIS and H3 data models. Open data (Overture Maps, OpenStreetMap) or your own GeoJSON and shapefiles. Built new or fixed in place, on your stack.
- Buy vs build, decided honestly — Mapbox, ArcGIS, CARTO, Felt, Geoapify, and Stadia Maps are the right answer for a lot of products. I'll tell you when they are, when a vendor bill or a data constraint says otherwise, and then build whichever side wins — including the custom PostGIS and tile pipeline when that's the call.
- Build the v1 — Turn an idea into a working product in production. Architecture, data model, backend, frontend, infra, deploy. Ruby on Rails, Go, or Python on Postgres. Real users, real load, real numbers.
- Cut through complexity — Sprawling system, tangled product, stalled migration, data nobody can untangle. I find the real problem, narrow the scope, and ship something clean in its place.
- Data pipelines and performance — Ingest, dedup, and serve large datasets with minimal infrastructure. SQL-first on Postgres, PostGIS, and DuckDB; AWS, DigitalOcean, and Cloudflare when they're the cheapest way to run it.
- Be your fractional senior engineer — A few days a week, async, no standups. Set the bar, ship the hard parts, hand off cleanly.
Stack I work in most: PostgreSQL, PostGIS, H3, Ruby on Rails, Go, Python, DuckDB, GeoParquet, GDAL, Planetiler / OpenMapTiles (Mapbox Vector Tiles), Overture Maps, OpenStreetMap, AWS, DigitalOcean, Cloudflare (Pages, Workers, R2).
Engagement Shapes
- Project build — Fixed scope, fixed end date, working software at the end.
- Sprint — One to three weeks. A working proof, not a deck.
- Embedded — A few days a week, ongoing, treated as part of the team, with a defined scope per month.
- Advisory + hands-on — System and data design, with me in the codebase when it matters. Cadence depends on the problem.
Sprints and project builds are scoped upfront, so the work is clear before it starts.
Who I'm a Fit For
- Small and mid-sized businesses with a map or data feature that matters — store locators, service areas, delivery zones, fleet or asset tracking, place search, geo-reporting — and no in-house specialist to own it.
- Operators with a product or system that has grown messier than the business needs it to be.
- Founders pre- or post-PMF who need a real builder on the v1 — not a team to manage.
- Companies between hires who want one person who ships, not another seat to fill.
- Teams with a hard, narrowly scoped problem — messy real-world data, scale, geospatial, performance.
Who I'm Not a Fit For
- Pure architecture roles where I don't write code.
- IC seats with no ownership.
- Layers of approval before anything ships.
- Maintenance with no room to simplify.
What You Get
- Someone who treats the work like an owner, not a contractor.
- Working software, fast. A POC same-day where it makes sense; production quality on the timeline we agree.
- Collaboration that matches your team. Async-first to keep your meeting load and costs down — written updates, demos, SQL, and data samples carry the work — with calls, pairing, or daily check-ins whenever they move things faster. (One engagement has run entirely in writing for two years and kept shipping; I flex to your cadence, not the reverse.)
- A clean handoff. Code, docs, and decisions in a state your team can actually run with.
- References from current and recent engagements available on request.
Start a Project
If you have a map feature to build, a system to simplify, or a data layer that needs to actually work — send a short note covering:
- The problem and what "done" looks like
- Rough timeline and any hard deadlines
- What's already in place (stack, data, team, repos)
A few sentences is enough. I'll write back with whether it's a fit, what an engagement would look like, and the next step.