Answers to the sensible questions.
What if I'm not sure which package to buy?
Two options. If you already know roughly what you want built, book a free 20-minute scoping call and we'll point you at the right package. If you want a full map of where AI could save you time and money across your whole business before committing, take the $950 AI Assessment. You'll get a custom 10 to 15 page report in 48 hours covering top pain points, tool recommendations, quick wins, and a financial impact estimate. The $950 is credited against any implementation package booked within 30 days.
What happens if the build doesn't work?
Every package has written acceptance criteria signed at kickoff. If the delivered build doesn't meet them, we fix, re-scope, or refund in line with the signed agreement. This is in the contract, not the pitch.
Who actually builds the thing?
A vetted senior AI operator matched to your package. Every operator has delivered paid AI work in that exact category, passed a practical build challenge, and cleared a reference check. Around 8 percent of applicants make it in.
What if my project is bigger than one package?
We stack packages or scope a custom fixed-price build. If what you need is genuinely a six-month engagement, we'll say so on the first call and point you elsewhere.
Do I own the code and the prompts?
Yes. You own everything: code, prompts, connections and logins. It is delivered into your systems and your accounts. If you ever want to leave, you leave with everything.
How do you keep prices this low?
Scope is tight, operators are senior, no account managers. We package work into repeatable shapes. When something is in a package, we've delivered it more than a dozen times.
Is there ongoing support after the build?
Each package includes a defined support window (7 to 30 days depending on scope). Longer retainers are available but opt-in, not default.
Can I talk to the operator before kickoff?
Yes. Every build starts with a 20-minute call between you and your matched operator. You confirm fit before any payment is made.
What regions do you serve?
Gateway AI is UK-based. We work with clients across the UK, EU and North America. Operators are distributed across those regions so timezone coverage is handled.
What tools and models do you use?
We default to the best tool for the job: Anthropic Claude for reasoning, OpenAI for specific vision tasks, Deepgram/Eleven for voice, and Make/n8n/Zapier for glue. We avoid lock-in and document everything we pick.
How is data handled?
Data ownership stays with you. We sign NDAs and DPAs as standard. Private builds (Knowledge Assistant especially) run in your own cloud tenant. No client data trains third-party models.