Companies are deploying AI agents by the thousand and cannot check them. We sell the check.
A re-runnable verdict on every agent run — priced like an audit, sold like a subscription. First enterprise contract above £1M, method validated independently, patent priority from 2025. We are raising to build sales.
Gartner expects agents in four in ten enterprise applications by end-2026, up from under one in twenty a year earlier — and forecasts that over 40 % of agent projects will be cancelled by 2027, naming inadequate risk controls as the cause. The EU AI Act's record-keeping duties, ISO/IEC 42001 and NIST AI RMF all presume evidence of what an agent did. Nobody can produce it for agents they did not build.
- 1Position. We sit beside the agent, not inside it — so the product does not depend on which model or runtime wins.
- 2Independence. The method is owned by a university consortium, coordinated by Edinburgh Napier. A model vendor can copy a check; it cannot confer independence on it.
- 3Data. Every verified run adds to the trace corpus and, from 2028, the signature database. UK patent priority 2025; two further applications in preparation.
The round funds sales, not research; the science is grant-funded through joint bids. Figures, model and use of funds are in the deck, sent on request.
Xplore Intelligence Ltd · Edinburgh · hello@xploreintelligence.co.uk · Stand: Innovation Alley, Tue 15 – Wed 16 Sep