Output validation

🔍 QA validation checklist

1. Logical consistency

• [ ] Do the cases make sense?
• [ ] Are there contradictions?

2. Real coverage

• [ ] Are critical scenarios missing?
• [ ] Did the AI ignore risks?

3. Business alignment

• [ ] Does it reflect real rules?
• [ ] Or does it invent logic?

4. Data quality

• [ ] Is the data realistic?
• [ ] Does it apply to the domain (insurance, fintech, etc.)?

5. Depth

• [ ] Are they superficial, or truly useful?