Security

Where your data goes, and who sees it.

The vendor questionnaire, answered before you send it. Everything on this page is true today. If something isn't listed, I don't claim it.

Data handling

Seven questions, plainly answered.

01

Where data lives

Live systems run in an account you own: your cloud provider, or a server in your company's name. We choose the region together.

My own infrastructure only holds audit mock-ups and demos, built on sample or anonymised data.

For the most sensitive documents, the AI can run on a local model, so the documents never reach an outside provider.

02

Who can see it

Me. I build and ship everything myself: no subcontractors, no juniors, no hand-offs.

I only access the data a process needs, and only while the work needs it.

03

AI providers and training

I mostly use Anthropic's Claude models, and OpenAI's where they fit the job, through their commercial API terms.

Under those terms, the data you send isn't used to train their models. If you'd rather no provider sees a document at all, a local model is the alternative.

04

Retention and deletion

At handover, or if we stop working together, I delete the client data and copies I hold and confirm it in writing.

I don't keep credentials after handover. The only records I keep are the contract and invoices, as the law requires.

05

Legal footing

I'm happy to sign an NDA before we start scoping.

Scope, price and support after launch are agreed in writing before any build starts.

06

Ownership

The code, the data and the accounts it runs on are yours at handover.

Everything is documented and built on mainstream tools, so another developer can pick it up.

07

Incidents

If I become aware of a security incident affecting your data, I tell you within 12 hours.

Have a security questionnaire?

Send it over. I'll answer it in writing.

Email ryan@algorithm.mu→