FAQ

Frequently asked questions

Pricing, ownership, security, evaluation and support — the questions teams ask us most.

What does a typical engagement look like?

Most clients start with a short discovery or a scoped pilot (usually 3–6 weeks) before a larger build. You get a fixed price and clear success metrics before any billable work begins.

Who owns the code and models?

You do. Source code, model artifacts, prompts and documentation are transferred to you on final payment, as set out in our Terms of Service.

Do you work with our data securely?

Yes. We work under an NDA and a data-processing agreement, use least-privilege access, and can run within your own cloud environment when required.

Can you use open-source models instead of an API?

Often, yes. We help you choose between hosted APIs and open-source models based on cost, latency, privacy and quality, and we can self-host where it makes sense.

How do you measure whether an AI feature actually works?

Every project ships with an evaluation set and metrics agreed up front. We report results honestly, including limitations and failure modes.

What does it cost?

Projects are scoped and quoted in writing before work starts. Pricing depends on the use case, data readiness and integration scope — there are no hidden fees.

Do you offer ongoing support?

Yes. We offer monthly support and MLOps retainers covering monitoring, model updates, evaluation and improvements, with response times in your agreement.

Didn't find your answer?

Ask us directly — we're happy to help.