Resources

Practical tools for figuring out whether an AI idea is useful, risky, overhyped, or just a very expensive way to avoid writing a process document.

AI Use Case Gut Check

Before you buy, build, or announce anything, ask:

1. What exact job are we trying to make easier?
2. What input does the AI need, and who owns that data?
3. What does a good output look like?
4. Who verifies the result?
5. What happens when it is wrong?
6. Is this automation, assistance, search, drafting, classification, or executive wishcasting?

Good AI projects tend to be boring

• Summarize long internal documents into decision briefs
• Draft standard responses for human review
• Classify incoming work and route it faster
• Extract structured data from messy text
• Generate checklists, test cases, and implementation plans
• Watch for anomalies and notify a human before the wheels exit the vehicle

Bad AI project warning signs

• “We need an AI strategy” but nobody can name the workflow.
• “It will save headcount” before anyone has measured the work.
• No one knows who checks the output.
• The demo only works with perfect inputs and a sales engineer nearby.
• Security review is treated like optional garnish.

More templates, scorecards, and checklists will land here as the site grows. Yes, a resource page with “resources coming soon” is legally required by the internet. I don’t make the rules.