Trust

What you can check before you trust us

We have no customers yet, so this page starts with what does not need one. Everything in the first four sections is on this site already and can be checked without asking us anything.

Complipath decides with rules, not a language model. Every classification is made by a rule engine and cites the article, the engine version and the law version. AI only drafts text from your own documents. Every sentence links to its source, and a person confirms it.

Twelve questions a reviewer would ask, answered in public

A reviewer wrote down twelve questions cold. The answers are published as given, and the answer to four of them is no. Seven carry a dated correction that keeps the wording it replaced, so you can see what we got wrong and when we fixed it.

Read the twelve questions.

We ran the engine on ourselves and published what it got wrong

We answered our own engine's questions about our own product, ran the classification and published the output verbatim — including the sentence it got wrong about us. A vendor's engine run on the vendor is the only test where the vendor cannot choose the input.

Read what it decided.

The legal texts we derive from, pinned by sha256

Every classification is derived from source files published here with their sha256 sums. If one of those files differs from what we pinned, our build fails. You can download them and check the hash yourself.

Check the corpus.

Every assessment records which version of the law it rested on

The Act has already been amended once. An assessment that cannot say which text it answered is a filing cabinet, not evidence. Each of ours carries the version of the law it was derived against, and the amending act is published here point by point.

Read the amending act.

If your security team is reviewing us, the questions it will ask — who sees which data, where it is stored, whether it trains AI models, how it is deleted, how people sign in — are answered in one place, as a page and as a PDF: the security review.

Hosting in the EU. Your workspace is stored in Frankfurt: the database in Supabase's eu-central-1 region and the app's functions in Vercel's fra1 region, read on 8 October 2026.
Audit log. On Starter and in the trial, Settings → Audit log shows who did what and when, and exports it as CSV. No member can edit or delete a line, an owner included; deleting the whole workspace removes it with everything else.
Two-step sign-in. Every plan can turn on a code from an authenticator app after sign-in, and the workspace owner can require it for everyone.
Where AI drafts are processed. Drafts and readings are written by Anthropic's language model. The processing region is global, the default in Anthropic's data residency documentation (read on 8 October 2026), and the app sets no other region on its calls; the subprocessor list says the same.