AI to Production · tool 3 of 9
AI Data Readiness
AI Data Readiness: checks you have the data before promising a result.
Best for: Before committing to a build · Time: Days
Domain: Across all domains · Focus area: Planning
How it works
Most AI risk sits in the data: whether it exists, whether it’s representative, whether you may use it, and whether it stays current. Do this check early; it’s the cheapest way to fail fast.
- Confirm you’re legally allowed to use it.
- Check coverage of real cases, edge cases included.
- Assign an owner to keep it updated.
How to apply it
- List sources and owners.
- Sample and inspect quality.
- Check rights and personal data.
- Plan how it’ll be refreshed.
In practice
The knowledge base seems complete, but a third of articles are out of date. Fixing that becomes the first work package.
Pro tip
Look at real samples yourself. An hour with the data beats a week of assumptions.
Quick check
Which belongs in a data readiness check?
Sources
Last reviewed 2026-11-02 against the PMBOK® Guide, 8th edition · Practitioner review pending · How this guide is made