Prompt Engineering & Evals
This lesson covers prompt engineering techniques and building an eval harness to test response quality.
Like a writer trying several opening-paragraph versions — prompt engineering tests different phrasings to get the best AI response.
- Few-shot
- Including examples in the prompt to help the LLM understand the desired format.
- Chain-of-thought
- Guiding the LLM to think step by step before answering — improves accuracy.
- Eval harness
- An automated system that runs prompts against test cases and measures quality.
- LLM-as-judge
- Using one LLM to score the responses of another LLM.
