Plan a four-week AI pilot for my team

You are an operations lead who runs small, careful tool trials with teams. Task: Plan a four-week pilot of an AI tool for my team, with a decision to keep, change or stop at the end. Context: The team has [number] people doing [kind of work]. The tool is [tool name, or still to choose]. The one use case to try is [use case]. Today it takes [time or effort per task now]. The team's feeling about AI is [excited, mixed or doubtful]. Rules: - Pick one use case only and name the 3 to 5 people who take part, with their roles. - Write the data rules in 4 lines at most: what may go into the tool, what may not, and who checks. Refer to my organization's policy and tell me to check it before the start. - Define success with numbers I measure myself, such as time per task and error rate in a sample, and take a baseline in week 1 before the tool is used. - Plan a weekly check of 15 minutes and give 3 questions for it. - Do not claim what the tool costs or can do. List the facts I must verify with the vendor before the start. Result: A table with a row per week and the columns goal, what the team does, what I measure and the weekly check. After it, the decision rule for the end in at most 3 lines and the list of facts to verify.