Every AI system and agent, on one record.
The registry is where each AI system and agent your organisation builds or buys is captured once — owner, purpose, data, third parties, lifecycle. Classification, documentation, questionnaire answers and your trust page all derive from it.
You can't govern what isn't on the record.
Most organisations cannot answer a simple question from their board, their auditor or an enterprise buyer: which AI systems are we actually running, who owns each one, what data do they touch, and which laws apply to them. The answers live in spreadsheets, ticketing systems, security reviews and individual engineers' heads. When a regulator asks, or a deal stalls on the AI questionnaire, the work begins from zero.
A checkbox in a GRC tool is not a control. A line in a spreadsheet is not a register. Governance owned operationally — by the people building and running each system — begins with a single, accurate record of what exists.
Capture each system once. Keep it current operationally.
A registry is only useful if it stays true. Hael's registry is structured so the people who build and run each AI system can keep it accurate as part of normal work — not as a quarterly compliance exercise.
The registry is the source. Everything else derives from it.
The registry is the source. Everything else — classification, documents, questionnaire answers, your trust page — is derived from it, and stays consistent because it shares that source.
Register a system once. Every framework it touches is mapped from that one record.
The same record satisfies overlapping obligations across the EU AI Act, ISO/IEC 42001, NIST AI RMF and the US state laws — map it once, and Hael reuses it everywhere it recurs.
Start with the record. The rest follows.
Bring one live AI system to the call. We'll register it with you and show what Hael derives from it — classification, the Annex IV file, the buyer-questionnaire answers, the trust-page entry.