- Step1: Install QueryCraft via pip: pip install querycraft
- Step2: Configure API keys for desired LLM providers
- Step3: Define prompt modules and pipelines using the QueryCraft SDK or YAML
- Step4: Set up evaluation metrics for response quality and cost
- Step5: Execute queries and collect results with the CLI or Python code
- Step6: Analyze results, compare prompt variations, and refine prompts iteratively