develop text based search Backend

[Nauman-S]

So this is how the UI is supposed to look like, maybe for starters in the chat section we can just return kNN docs, but in an ideal scenario the user should get a response of the relevant information in text based (extension)

This is frontend to backend connectivity - ChatGPT and other models use HTTP + SSE

Spring provides Spring AI

Video:
[ChatClient to provide rag type models out of the box](https://www.youtube.com/watch?v=umKbaXsiCOY&t=1415s)

Documentation:
Spring AI - https://spring.io/projects/spring-ai
Spring Embedding Creation - https://docs.spring.io/spring-ai/reference/api/embeddings.html

Option 1: Chat Stuffing
Chat stuffing is just concatenting the context kNN docs as string to the user request

• ISSUE: MODELS CAN ONLY INGEST SO MUCH CONTEXTUAL DATA + PRICING OF HAVING A TON OF TOKENS
  Option 2:

• Create a backend endpoint that streams dataa

• Tokenization of user text

• Tracking of token usage

• Creation of embeddings of user text

• User Authentication