Project · An AI opportunity assessment for your product

Produce a decision document, not code: map at least five candidate AI features for a real (or sample) product onto impact/effort, score each through the go/no-go rubric, run a one-day feasibility spike on your top candidate against real examples, and deliver a prioritized recommendation with the three taxes quantified.

• An opportunity map of 5+ candidate features scored on impact and effort.
• A go/no-go rubric score for each candidate with weak dimensions called out.
• A spike result (pass rate on real examples) for the top candidate.
• A one-page recommendation with cost, latency, and trust quantified for the chosen feature.

• Each candidate is framed as a job-to-be-done with a measurable outcome, not 'add AI'.
• At least one candidate is correctly ruled out as a plain-code or low-value task.
• The spike uses real product examples and reports an honest pass rate.
• The recommendation puts concrete numbers on cost-per-use, p95 latency, and error tolerance.