RAG ⚐

up: LLMs ⚐ Retrieval Augmented Generation. This is best to provide a knowledge base (as opposed to Finetuning).

• LlamaCloud
• relevance.ai
• Hosted Vector Databases
• LlamaIndex

Set up RAG

1. Vectorise the source data into a Vector Store (RAG)
   • I think LlamaIndex can do that with the CSV loader
2. Do a similarity search ()
   • Get the relevant docs
3. Pass docs to LLM and Prompts
   • Email RAG Prompt
4. Execute RAG

Terms

• Vector Store (RAG)
• Retriever
• Retrieval Question-Answering Chain