Most AI looks great in a demo and falls apart in production. We're the two engineers who fix that — building systems that hold up when real users show up.
Start a conversationWe're not an agency scaling headcount. We're two engineers who've spent years in the trenches — Rails, GCP, LLMs, the full stack — and now we channel all of that into making AI work for real businesses.
We're heads-down on our own projects right now and not taking clients. But if you're building something interesting with AI and want to think out loud — we're always up for a good conversation.
You don't need AI everywhere — you need it in the right places. We figure out where it actually helps your business, then design the system around that truth.
StrategyThe difference between a cool demo and a viable product is often latency and cost. We handle quantization, caching, batching — the boring stuff that makes AI economically real.
OptimizationLLMs are only as good as the context you feed them. We build retrieval pipelines that ground answers in your data — accurate, fast, and without the hallucination tax.
DataGetting from "it works on my laptop" to "it handles 10k requests a minute" is where most AI projects die. That gap is exactly where we live.
EngineeringThe best AI disappears into the product. You stop noticing it and start relying on it. That's the bar we aim for.— How we think about our work
Not taking new clients yet — but our inbox is always open for good conversations about hard AI problems.
Say helloWe usually reply within a day.