Table of Contents

• Loom Video Questions
• Introduction
• Evaluation Framework
• Critical Success Factors
• Evaluation Categories
  1. Prompt Engineering Quality (20%)
  2. Strategic Use of AI Tools, Model selection, and Cost-Efficiency (20%)
  3. Output Consistency, Reliability, and Content Quality (15%)
  4. AI for Data Classification and Categorization (10%)
  5. AI Output Validation and Error Handling (15%)
  6. Context Management and Data Synthesis (10%)
  7. Creative Problem-Solving with AI (5%)
  8. Documentation and Prompt Transparency (5%)
• Best Practices
• Common Pitfalls to Avoid
• Claygent Model Selection Guidelines
• Project Examples

Loom Video Questions

1. Walk me through your Claygent prompt. Why did you structure it this way?
2. Show me an example where Claygent returned low quality, incorrect, or incomplete data. How did you catch it? What's your validation logic?
3. Walk through a scenario where Claygent can't find the information you asked for. What's your fallback?
4. Why did you choose Claygent for [insert research task] instead of a standard enrichment provider?
5. Why did you choose X AI model for this task?
6. Show me how you're using AI to categorize or score something. What categories or scoring criteria did you give it? Why those specifically?
7. Show me how you're using AI for data extraction vs. data generation. What's the difference in how you prompt for each?
8. Show me your prompt iteration history. How did you improve it from version 1 to the current version?
9. Show me how you would edit your Claygent prompt to extract data in a different format (e.g., from paragraph to bullet points, or vice versa).