Frequently asked questions.

The production audit is $7,500 fixed and takes a week. Builds start at $30,000, priced module-by-module in the spec the audit produces. Private LLM deployments start at $45,000 (readiness assessment $9,500). Maintain starts at $1,500/month. Every price is fixed before we start: no day rates, no surprises.

Yes, that’s a core service. We take it over rather than start again: keep everything that deserves to live, fix what doesn’t. It starts with the audit: security, data model, infrastructure, and a fixed-price spec to get it production-ready. And no, we won’t laugh at the code.

Usually, yes. Most pilots stall on integration, data plumbing and governance: engineering problems, not model problems. The audit tells you the honest distance to production; the spec prices the journey; you decide with numbers in hand.

The audit produces a written specification: modules, each with acceptance criteria and its own fixed price. That spec is the scope. Change your mind later? The spec gets a revision, with a new price on the changed part only. You’re never billed for drift.

Open-weight models (Llama, Mistral, Qwen and friends) running in your own cloud or on your own hardware, instead of sending your data to a third-party API. You’d want one if your data is sensitive, regulated, or proprietary, or if your API bill has become a line item the CFO asks about. We assess readiness first ($9,500), including whether you’d be better off not doing it.

Monthly: uptime and error monitoring with a written report, security patching and dependency updates (logged), backup and restore tests, and LLM model migrations when providers retire models, which they do on their timetable regardless of yours. Higher tiers add eval runs on every model change, improvement days, and fractional delivery leadership. Our founder is a former CISSP and Certified Ethical Hacker, so the security patching is done by someone who knows what they’re patching.

You do. 100%, including the specification. Take it to another builder tomorrow if you like. We’d rather earn the next project than trap you in this one. Read our terms →

Yes. It’s where we came from and it’s still the fastest way to test an idea. But a prototype is where the pipeline starts; it was never the finished product. Ours come with a straight answer about what production will take.

Constantly. You’re most of our rescue work. You built something real with AI tools; it grew users; now it needs an engineer. No specs required, no jargon, no judgement.

Audit report in one week. Most builds ship in 4–10 weeks depending on the spec. If the audit says your app is closer to production than you feared, we’ll say so: sometimes the honest answer is a short fix list rather than a big engagement.

Yes. We work in the open: you see progress every few days, tested against the spec as we go, with nothing held back until the end. See how it works →

Between modules, yes: each module in the spec is independently completable, so you’re never stranded mid-air. Maintain is monthly rolling; leave whenever, take everything with you.