Your team
Features and ideas, building with the AI tools they already use, and deciding what the app does next.
The App Foundation adds security, logins, backups and monitoring to apps your team has built with AI tools like Lovable, Replit or Claude. We put these in underneath the app, then hand it back so your team can keep building.
Any one of these is a reason to review the app.
The app can only reach the data its job needs, with no path to take payments, change billing or delete records. Keys are stored on the server, not in the code.
Each person signs in individually with an email link. Access is set by role, so managers see financials and staff see what they need.
A record of who changed what and when, available at any time.
Version history, a test copy before changes go live, one-step rollback, and rules that stop AI tools breaking existing features.
Failures alert us straight away, and there's a switch to stop the app quickly if needed.
A working session on building safely (try it, test it, then put it live) and a short written runbook.
Features and ideas, building with the AI tools they already use, and deciding what the app does next.
Security, logins and permissions, data connections and backups, and keeping the app running and up to date.
01
App Safety Check
About a week. We review the app with whoever built it and provide a plain-English report of risks, with a fixed price for the Foundation. A small fixed fee, credited in full if you go ahead.
02
The Foundation
3 to 5 weeks. Everything listed above, built and tested on a copy, then switched on with your team, at the fixed price set by the Safety Check.
03
Keep-alive (optional)
A flat monthly fee covering monitoring, security updates and fixes to anything we built, plus set hours each month for changes or reviews.
Each Foundation is scoped to the app. The main factors are how many systems it connects to, how sensitive the data is, and how many people use it.
Used by 5 staff. Reads from one system. No customer payments.
Used across two sites. Reads and writes to the member system and a spreadsheet tool.
Customers log in. Holds personal and payment-related data.
The code, data and accounts are yours from day one. Hosting, database and AI usage are billed to you directly, at cost, with no markup. Anything outside the agreed scope is estimated and approved before work starts.
Other questions can be covered on a free call.
Vibe-coded apps built with tools such as Lovable, Replit or Claude, and most apps that use a standard database. The Safety Check confirms yours.
No. We add what's missing underneath and leave your team's work in place. Any required changes are explained first.
No. We work on a copy and switch it on with your team once it's tested.
You do, from day one. Running costs are billed to you directly, at cost.
Yes. The report is yours to act on with anyone, including your own team.
About a week for the Safety Check, then 3 to 5 weeks for the Foundation.
Instructor-led workshops on Claude, ChatGPT or Microsoft Copilot, for teams, groups and individuals.
Find out more →SetupA two-hour live session that configures Claude for your business, including your AI Operating System.
Find out more →BuildCustom AI tools and automations, scoped and priced through a paid discovery.
Find out more →A 30-minute call to discuss your business and whether this service suits it.
Book a free strategy call30 minutes · Free · No obligation