← FDE Playbook
Free · one pageUpdated 2026-10-07
AI Discovery Cheat Sheet
Business discovery with the sponsor and users
- What decision or task should get better, and for whom?
- How is it done today, step by step, and how long does it take?
- How will we know it worked? What number, measured how, today's baseline?
- Who owns that number and will sign off on success?
- What happens if the AI is wrong? Who is affected and how badly?
- What has been tried before, and why did it stop?
- What is the deadline, and what is driving it?
Technical discovery with data, security and IT owners
- Which systems hold the data, and who can grant access?
- How is access controlled today (identity provider, groups, roles)?
- What regulation applies (privacy, health, finance, sector rules)?
- Where may data be processed and stored? Any residency rules?
- Which external AI services are already approved, if any?
- Where must the result appear: which app, channel or workflow?
- What logging and audit does security expect?
Red flags that should pause the project
- Nobody can name a measurable outcome or owner.
- The data needed does not exist, or nobody can grant access to it.
- The AI would make an irreversible or regulated decision with no human review.
- Security has not been told about the project.
- Users were not consulted and the sponsor cannot name one.
Send the same day
- A one-paragraph summary of the problem in their words.
- The success metric and baseline you heard, for them to correct.
- Open questions with a named owner for each.
- The proposed next step and date.