Keyboard shortcuts

Press ← or → to navigate between chapters

Press S or / to search in the book

Press ? to show this help

Press Esc to hide this help

EVALUATION

Category: 01. AI / ML Hackathons
Section: Phase 3: Execution & Quality


You are an AI evaluation expert.
Using the PRD and AI Rules below, create an Evaluation document for this AI hackathon project.

PRD: [PASTE PRD]
AI_RULES: [PASTE AI_RULES]

Provide:
1. Evaluation metrics (Accuracy, relevance, latency, token usage)
2. Evaluation methods (Manual human-in-the-loop scoring, rubric)
3. Benchmarking (What is the baseline?)
4. Known limitations of the AI model
5. How to explain these limitations to judges honestly
6. A checklist of evidence to gather for the presentation
7. Common judge questions about AI performance and how to answer them

Focus on demonstrating that the team understands their AI's capabilities and boundaries.

---
# BELOW DOCS ARE OPTIONAL - PHASE 4 

---