Enterprise pilot programme
Prove AI governance in eight weeks, on real obligations.
A defined investment, measurable outcomes and a controlled path from evaluation to production — delivered on working software, not a slide pack.
- Fixed fee
- Quoted per pilot in the Order Form
- 8 weeks
- From agreed kick-off, after mobilisation
- 15
- AI assets and use cases registered
- 30 days
- Warranty and hypercare after handover
What you receive
A working control plane, configured to your programme.
The pilot covers one agreed business unit or governance programme, end to end, so the organisation can make a defensible decision on production adoption.
- A working, configured AIRAS environment for one agreed business unit or governance programme.
- Up to 15 AI assets and use cases registered, with up to eight structured assessments completed.
- Three selected use cases taken through the full review, evidence and governance-decision pathway.
- One standard integration proof delivered, with the architecture defined for up to three future integrations.
- Role-based training, operating procedures, executive outcome reporting and a production-conversion roadmap.
Delivery plan
Four controlled phases across eight weeks.
- Weeks 1–2
Mobilisation and configuration
Scope confirmation, tenant provisioning, role and permission model, policy pack selection and ruleset configuration under governed change control.
- Weeks 3–4
Register and assessment
AI systems, models, agents and embedded vendor features registered with named owners, then assessed against the deterministic, versioned risk model.
- Weeks 5–6
Review, evidence and decision
Reviewer routing with segregation of duties, control assignment, hashed evidence capture and versioned decisions with conditions.
- Weeks 7–8
Monitoring, reporting and conversion
Monitoring and incident workflow, executive reporting, exportable assessment and decision packs, and a costed production-conversion roadmap.
Assurance evidence
Evidence mapped to the obligations you are asked about.
Each pilot produces records against the frameworks your regulators, auditors and customers already reference.
| Framework or instrument | Pilot evidence produced |
|---|---|
| EU Artificial Intelligence Act | AI inventory, role and accountability mapping, risk classification, human oversight, technical and evidence records, incidents and lifecycle monitoring. |
| GDPR and automated decision-making | Data inventory, privacy review linkage, decision scope, human oversight, notices, and review and challenge evidence. |
| NIS2 | Identity, least privilege, tenant isolation, supplier assurance, incident workflow, audit and recovery evidence. |
| DORA | Financial-services pack, operational resilience, integration and vendor records, incidents and control evidence. |
| ISO/IEC 42001:2023 | Governance roles, policies, risk treatment, objectives, competence and continual-improvement records for an AI management system. |
| NIST AI RMF and GenAI Profile | Govern, Map, Measure and Manage activities evidenced against assessed context, with GenAI-specific risk actions. |
| OWASP GenAI and agentic guidance | Prompt, retrieval, output and action-control patterns, tool allowlists, privilege boundaries and supply-chain checks. |
Framework references describe design alignment and evidence support. They do not represent certification, legal advice or a guarantee of compliance.
Data position
Synthetic, de-identified or expressly approved pilot data only. Personal or sensitive production data requires prior security, privacy and contractual approval.
Commercial status
Non-binding until the parties execute an Order Form or Services Agreement. Scope, contracting entity, tax position and data-processing terms are confirmed before mobilisation.
Reference AIRAS-PILOT-2026-001
Scope your controlled pilot
Tell us the business unit, the obligations in scope and the decision you need to reach. We will return a tailored pilot proposal and delivery plan.
Commercially available now. Pricing is quoted per pilot and confirmed in an executed Order Form.