# Copyable starter questionnaire

> **1. Value and owner:** What business result are you trying to improve, how will you recognize useful progress, and who owns that result?
>
> **2. Approach:** Are you buying, building, enabling AI already embedded in a product, combining those approaches, or still deciding?
>
> **3. AI and architecture:** What tools, models, providers, and major components are known? What remains unknown? Add a rough sketch if one exists.
>
> **4. Data:** What information might go in or come out, where does it come from, how sensitive might it be, where might it go, and what is known about reuse or retention?
>
> **5. People and environment:** Who would use it, who could be affected, where would it run, and how would users get access?
>
> **6. Connections and actions:** What could it read, recommend, create, send, or change? Which integrations are contemplated, and where would a person review the result?
>
> **7. Consequences:** What could plausibly harm a person, customer, employee, business process, obligation, or important dependency if the AI is wrong, misused, manipulated, unavailable, or trusted too much?
>
> **8. Next experiment:** What is the smallest useful thing you want to try next, and is any version already operating?
>
> **If buying or enabling embedded AI:** Which provider and service are under consideration? What settings, data terms, retention, access controls, logs, change notices, and exit options are visible today?
>
> **If building or combining components:** Which model, retrieval source, tool, hosting environment, service identity, and evaluation method are contemplated? Which parts remain open choices?