Software engineering for San Francisco.
Senior, embedded engineering for San Francisco-area teams shipping AI, machine learning, and custom software in production. Pick a practice below — or talk to us about something we haven't listed yet.
San Francisco is where generative AI hype is loudest, which is exactly why the quiet part matters most here: architecture, evals, and cost engineering that turn a Series A demo into a system that doesn't melt under real traffic. Forty-plus GenAI projects shipped across ten-plus model providers, for a client base that includes multiple YC company alumni.
Full Pacific Time overlap means pair programming and design crits happen without a time-zone tax, and our team has been the in-house team before — we know what runway feels like and won't gold-plate a system your Series A budget can't carry.
Fitness Flywheel is a good look at how we approach consumer AI product work: an adaptive coach engine that reshapes workouts and nutrition guidance in real time, shipped across iOS and Android. See how we built it →
Do you work with pre-seed or seed-stage startups?
Yes, on focused scopes — and we'll say so directly if your stage is better served by a senior in-house hire than by an agency.
Can you help us choose which model to build on?
That's often the first deliverable — we benchmark on your data, against your latency and cost constraints, and write down why.
Do you take equity as part of payment?
Cash-only by default, though we've done blended cash-plus-equity arrangements for early-stage clients we believe in.
How fast can you ship a working prototype?
Two to four weeks for a serious prototype. We won't ship a fake one faster — it costs more to unwind later.
Tell us what
you’re building. We’ll ship it.
Send a few details and a senior engineer — not a sales rep — gets back to you with a clear next step within a day. In a hurry? .