I have some data in Aurora in different tables, I want to perform RAG for it.
For that I to make a micro service(Lambda) that can generate embeddings of different table and store that data in Aurora.But the vectors upsert should be based on some database stream collection.
For Example : If there are 5+ database updates , then I should regenerate embeddings and I should have a endpoint where I can send user Id to sync data with backend(regenerate embeddings).
How can I do it.
I tried using Faiss and dynamoDB initially but problem is that the backend is shifting to Aurora and Faiss is increasing Latency.
Aryan Raj is a new contributor to this site. Take care in asking for clarification, commenting, and answering.
Check out our Code of Conduct.