Hello,
Recently, I've been working on creating contract documents, scoring, and conversational apps using generative AI.<p>I'm not only interested in RAG but also want to improve the accuracy of prompts. I'm currently using around 100-200 prompts and would be happy if I could do component-based prompt engineering.<p>I am comprehensively using:<p>1. OpenAI Eval
2. Azure Prompt flow
3. Promptfoo
4. LangSmith
5. Wandb<p>If there are any better hacks, I would like to know.