AI CONSULTING & AUTOMATION / A PRACTICAL ANSWER
What data questions should you ask before an AI project?
The short answer
Before an AI project, establish what information the workflow needs, who may access it and how the proposed systems would process it. Data readiness includes relevance, quality, permissions and retention decisions. Do not assume that every available document is appropriate to send to a tool or expose through an assistant.
What to establish
Map the sources, intended uses, destinations and people responsible for approval. Identify personal, confidential or restricted information and seek suitable specialist review where needed. Check current provider terms and settings for the proposed account and workflow rather than relying on a general product reputation.
How to assess readiness
Choose test material that can be used responsibly and keep access proportionate to the pilot. Plan updates, deletion, incident handling and the records needed for accountability. This preparation does not itself establish legal compliance or remove the need for a qualified review.
ILLUSTRATIVE EXAMPLE / NOT A CLIENT RESULT
Make it concrete.
A public product catalogue and a confidential client file may require very different handling. A pilot can often explore the workflow using prepared examples while the business resolves whether operational data belongs in the proposed setup.
Before you commission
- Map information, permissions and processing destinations.
- Verify current provider and specialist requirements.
- Plan proportionate testing, updates and incident handling.
A plain-text worksheet. Downloading does not send us your details.
Further reading
FROM THE QUESTION TO THE WORK
A useful next step.
Bring the decision, the existing situation and the information you have. We can help define a scope before you commit to a wider project.
Explore the relevant service