Brightwater chooses its first three AI projects (fictional)
Brightwater’s HR team of five is stretched. The HR director asks each team member for one AI idea and receives six. The team scores them together using Value + Feasibility − Risk.
| Proposal | V | F | R | Score Notes | |
|---|---|---|---|---|---|
| A. AI-assisted job descriptions and postings | 4 | 5 | 1 | 8 Frequent task, easy to review. | |
| B. Onboarding question assistant | 4 | 4 | 2 | 6 Needs clean policy documents. | |
| C. AI ranking of warehouse applicants | 5 | 3 | 5 | 3 Risk gate: legal review before any pilot. | |
| D. Theming of survey comments | 3 | 4 | 2 | 5 Once or twice a year. | |
| E. New-hire kickoff automation 4 | 4 | 2 | 6 | Clear rules; needs IT approval. | |
| F. Sentiment monitoring of chat 2 messages | 2 | 5 | −1 | Rejected: trust and legal concerns. |
Decision. The team picks A, B and E. They are spread across recruiting and onboarding, sit on the low and medium rungs, and each has an obvious measure of success. D is scheduled for the next survey. C scored highest on raw value, which is why it is tempting, but the team records that it will not proceed without legal review, bias testing and a plan for candidate notice. F is dropped.
What to notice. The highest-value idea was not the best first project. Scoring made the trade-off visible and gave the director a defensible reason for the choice.
Jot down key ideas in your own words. Saves automatically.