Don't take our word for it.

What we build

Aqta builds what an AI vendor hands the person who has to review its work. Seal owns that path for one live workflow: a policy check on the call path, a signed record of the decision event. The record can travel with an evidence pack and be checked with a published key. Without defensible evidence, consequential AI workflows stall in security, risk or buyer review. Seal gives the reviewer a record they can check independently.

What we claim, and what we do not

Seal signs determinations whose inputs, rules and outputs can be independently checked. It does not sign judgements that depend on reasoning the reviewer cannot see. Where a determination is made by somebody else, a measurement engine or a named reviewer, we carry it and record who made it, rather than producing a judgement of our own and signing that.

Three limits belong beside that claim, because stating them is what makes it credible. A signature does not establish that the record matches the computation that actually ran. It does not establish that every decision was recorded, since a decision never written down leaves no trace of not having been written down. And it does not establish that the decision was correct, fair, or lawful. Those need other evidence and other people.

How we operate

Aqta Technologies Limited is an Irish software company building a production gateway, an open attestation format, and published verifiers. Seal is our production focus.

The gateway signs on EU infrastructure. An evidence pack is available for vendor risk assessment on engagement.

We work from the review question backwards: what the buyer must check, which decision path can answer it, and how the resulting record remains useful outside Aqta.

How we are set up

Aqta is founder-led, and the signing gateway, the open attestation format and both reference verifiers were built in-house. Because the protocol and the implementation are developed together, both can be published as open artefacts and evolve together without waiting on each other.

Specialist advisers sit alongside that work rather than inside it: a managing director in risk consulting at a Big Four firm on model risk, counsel on the EU AI Act and privacy, and corporate counsel on AI-agent liability. They advise, they do not hold office. We name them in investor and partner material, with their consent, rather than on a public page.

Senior engineering hiring opens with the current raise, and open roles are listed on careers.

Focused work

Seal is the production focus. Bounds Pro applies the same approach to sensitive documents. Other research stays in Aqta Labs, separate from production buyer work.

Production

• 
  Seal at app.aqta.ai. Signed records for one live AI workflow, with a public format and independent verifier.
• 
  Bounds Pro at bounds.pro. On-device document redaction for clinical-narrative PHI. Documents never leave the browser. Engine open-source on npm, Apache 2.0.

See Aqta Labs for earlier-stage work.

Recognition

Listed in the EU Women in Digital best-practice repository (Certified partner tag; case study, not product certification). Organisation profile in the FIDES Ecosystem Explorer (EUDI wallet ecosystem catalog; a listing, not an endorsement), where our wallet-authenticated review of AI decision evidence is a published use case. Selected for InterTradeIreland Shared Island (WeBuild) with KPMG mentorship. Community partner of Hack Apertus, an official pre-event of the Geneva AI Summit 2027.

Attestation format: open ATTESTATION-v1 spec.

Aligned with UN Sustainable Development Goal target 16.6, “develop effective, accountable and transparent institutions at all levels”. We state that alignment; the UN does not endorse us, and it is the only goal we claim.

Company