We use a few strictly-necessary cookies, no marketing trackers, and cookieless analytics. Read our cookie policy, or pick a preference.
Four products: the AI Intelligence Centre, the Integrity Engine, AI Literacy, and Practice Workspaces. Each module exists because a regulation requires an answer, and each traces to the Article or Annex that demands it. Nothing here is generic compliance documentation for its own sake.
The governance knowledge graph. Classify systems, run guided discovery, score gaps, and query the corpus in natural language, with every answer cited and traceable to source.
Continuous post-market monitoring, held as a tamper-evident chain of custody. Every event is sealed, so one record serves the regulator, the auditor and a liability claim alike.
Where controls and evidence are captured as teams ship their own AI. Guardrails in the build pipeline, evidence generated as you work, without slowing delivery.
The mandated duty that lands the account and seeds the method. A role-based curriculum bound to the obligations each role owns, with per-seat attestation. It is where most teams start, and the layer the three products above build on.
Capture it once, prove it everywhere. Checklists map a control to a named act; we map at the level of the obligation, so one body of evidence satisfies many regimes and the divergences surface before they bite.
A guided evaluation to a definitive risk tier, checked against the full prohibited-practices catalogue.
Live1,272 atomic questions, one ask to one answer, routed to the responsible domain with evidence attached.
LiveCorpus-grounded answers with two-stage citation validation across 644 instruments.
LiveReal-time scoring with heat-map visualisation and prioritised remediation guidance.
BetaDrift detection, reclassification triggers, and incident reporting on each authority's clock.
In developmentRole-based certification bound to the obligations each role owns, with per-seat attestation.
LiveShip status is indicative and binds to each module's record · article references confirmed against consolidated text before publication
TRACE is how we turn a regulation into evidence you can hand a regulator. Five steps, run once across every regime, not once per regulator.
The AI systems you run and the obligations that attach to each.
Each system by risk tier against the Act's criteria.
The controls and documentation each tier requires.
The operations that keep those controls live.
The result, so the file stands up when someone asks.
Source: SI obligation corpus 2026-07-18 · research-grade, internally audited (29 pass / 0 fail) · not counsel-attested