AI PROJECT EVIDENCE CHECKLIST — KIF
https://trykif.com/guides/linkedin-alternatives-for-developers

Project name:
Dates:
Public link (only if I have permission to share it):

1. PROBLEM
Who used it? What did they need? What was the previous approach?

2. OWNERSHIP
What did I personally design, implement, evaluate, deploy, or maintain?
Which work belongs to teammates or an existing library?

3. APPROACH
Why this model or retrieval approach? What alternatives did I consider?
What were the constraints: cost, latency, privacy, quality, or infrastructure?

4. EVALUATION
What did I measure? What was the baseline? How was the dataset chosen?
What failed? Which improvements were measured, and which are observations?

5. PRODUCTION
Was this a prototype or deployed service? Who owned monitoring and incidents?
How were poor outputs, access control, and changing data handled?

6. EVIDENCE
What can I share: a public repository, approved demo, report, or reference?
Do not disclose private customer records, credentials, or employer information.
Do not invent metrics. An honest limitation is useful context.

Three-sentence summary:
We needed...
I was responsible for...
The result was..., supported by...