The problem
EU AI Act work is stuck in Word & Excel. Each new system feels like a fresh project, even when the obligations are the same.
Today
- Every AI system becomes a custom Annex IV / FRIA / PMS project with new spreadsheets and templates.
- There is no single governance engine behind the advice, making it hard to scale across systems and clients.
- Auditors and boards want repeatable processes, evidence and traceability – not just documents emailed as attachments.
The solution
EU AI Copilot turns assessments into Annex IV-style packs, readiness scores and a traceability matrix – in your own environment.
What EU AI Copilot does
- Runs a guided assessment and classifies risk (Art. 6 + Annex III).
- Generates Annex IV-style technical docs, FRIA, PMS plan, QMS brief and Instructions for Use.
- Computes GREEN / YELLOW / RED readiness and blocks “clean” exports when duties are missing.
- Builds a Control-Roadmap.csv linking articles → policies → tests → evidence → owner.
- Logs human sign-offs (attestations) in an append-only evidence ledger.
Pilot tiers (one-off)
Fixed-scope pilots using 1–5 “lighthouse” systems to prove the workflow on real AI deployments.
Tier 1 – Single System Lighthouse
Indicative: €4,500 (design partners from ~€2,900)
Tier 2 – Lighthouse Portfolio (up to 3 systems)
Indicative: €11,500 (design partners from ~€7,500)
Tier 3 – Portfolio Starter (up to 5 systems)
Indicative: €24,000 (extra systems from ~€3,000 each)
Deployment & security
- Managed Secure Sandbox – dedicated single-tenant instance in EU cloud, quick start, data deleted or handed over at end of pilot.
- Self-Hosted – runs entirely inside your infrastructure (Docker); local-first storage, no client data sent back.
Para despachos y equipos en España
El piloto se adapta a despachos boutique y equipos in-house que quieren
pasar de proyectos manuales a un modelo de gobernanza del AI Act más
repetible. Podemos trabajar en inglés y/o en castellano, y desplegar EU
AI Copilot en sandbox gestionado o en vuestra propia infraestructura.