AI Readiness Checklist
Seven dimensions that determine whether an AI pilot will reach production, and what to validate before your first pilot begins.
Most AI pilot failures aren't caused by model quality. They're caused by organizational conditions that were never assessed: unclear value, unmapped workflows, no business owner, inaccessible data, or missing governance. This checklist helps you identify those gaps before they become expensive.
Business value clarity
A specific, measurable outcome has been identified, not a category like 'improve operations', but a defined business result with a before-state that can be measured.
- Specific business outcome defined
- Before-state is measurable (cost, time, error rate, throughput)
- A business sponsor is named and accountable
Workflow clarity
The workflow that AI will change has been mapped at task level. The exact integration point, inputs, outputs, and exception paths are understood before implementation begins.
- Workflow mapped step by step
- AI integration point identified with clear inputs and outputs
- Human-in-the-loop touchpoints and exception paths defined
Ownership and accountability
A business function owns the changed workflow and is committed to operating it. AI programs without clear operational ownership rarely survive the transition from pilot to production.
- Business owner identified (not only a technical owner)
- Roles and responsibilities mapped for the changed workflow
- Decision rights established for AI output review and override
Data access
The data required for the AI system to function is accessible, of reasonable quality, and free of significant security or compliance constraints that would block the pilot.
- Required data identified and accessible
- Data quality issues assessed
- Security and compliance constraints reviewed
Platform readiness
The technical environment can support deployment, monitoring, and rollback. A pilot with no deployment path is a demo, not a pilot.
- Deployment environment identified
- Monitoring and output quality tracking defined
- Rollback plan in place
- Path from pilot to production is understood
Governance
Governance isn't a compliance layer added after deployment. It's the operating structure that determines whether a successful pilot can cross into production.
- Stage-gate criteria defined for production deployment
- Exception handling and escalation paths established
- Model update and change management process defined
Baseline metrics
The before-state must be measurable for the after-state to be demonstrable. Baseline metrics established before deployment are the foundation of any credible ROI case.
- Baseline metrics captured before deployment begins
- Success criteria defined in business terms
- Measurement method agreed with the business owner
When to proceed, when to pause
Proceed with a pilot when:
- Value and business outcome are clearly defined
- The workflow has been mapped at task level
- A business owner is named and committed
- Required data is accessible
- A basic deployment and rollback plan exists
Pause and do foundation work when:
- Value or outcome is unclear or contested
- The workflow has not been mapped
- No business owner is identified
- Data access requires significant engineering work
- Governance and accountability are undecided
Related resources
AI Strategy Consulting
Readiness assessment, opportunity mapping, and first-wave prioritization.
Learn moreAI Governance Consulting
Decision rights, review loops, baseline metrics, and responsible scaling.
Learn moreReadiness Assessment in Depth
Expanded guidance on each dimension with practical diagnostic questions.
Learn more