Central government
Departments and agencies improving services and casework.
Sectors
Government, councils, housing providers, charities and regulated organisations that want to use AI carefully.
Who we work with
Departments and agencies improving services and casework.
Councils improving resident services, social care, planning and revenues.
Resident engagement, repairs, tenancy and regulatory reporting.
Organisations doing more with limited budgets. Our advisory board includes experienced charity trustees.
Admin automation and online services that free up clinical time.
Banks and financial firms, where our founder spent much of her career.
Standards
We design every project against these frameworks. They are design commitments, not certifications. Our Trust Centre covers certifications and registrations.
From understanding users to running a reliable service
Open standards, cloud first, secure by design, no lock-in
The ten principles for using AI safely and well
Transparency records drafted for the algorithmic tools we deliver
Secure design, build, deployment and operation of AI
Applied to the services we host
Tested with automated tools and assistive technology
Privacy by design, DPIAs, data minimisation and data-subject rights
Controls for prompt injection, data leakage and excessive agency
Registered on the Central Digital Platform
Content and interfaces that can support Welsh
How to buy
We are a UK SME registered on the Central Digital Platform.
Questions buyers ask
No. OpenAI, Anthropic, Google and others spend billions on models that improve every few months. We put those models to work on your problems, and we design every system so you can switch to whichever model is the best value.
Wherever your requirements say. By default we use UK or European cloud regions, keep data in your own cloud account where possible, and choose model providers whose terms rule out training on your data. Every subprocessor is listed in our data protection agreement.
With the right controls, yes. A senior member of our team reviews every change, and each one is covered by automated tests, scanned for security and accessibility problems, and linked to a requirement.
Through direct awards below threshold, competitive tenders on Find a Tender and Contracts Finder, and frameworks as we join them. As a small company we can usually move quickly from first conversation to contract.
Several AI agents, each with a role such as analyst, builder or tester, work on each task. Senior people set the direction and approve every output. The work is more thorough and quicker, with a full record of who did what.
Yes. We often work as a mixed team, and we leave documentation, runbooks and skills behind so your people can run and extend what we build.