hello folks
i hav an idea and i want start to build it but before i have question based on the nature of the project and data
what should use to bulid it ?
when the data is static and its contains 50K document’s ,
should i use Chatgpt Api ?
or Langchain ?
or lamaindex ?
When you say “build”, what kind of development background do you have? Have you ever built a Saas application before? If so, then you may be able to leverage your existing experience and programming language skills to achieve your goal – you only need learn how to use the OpenAI APIs.
If you do not have this background, then Langchain seems like a safe bet as they have libraries pre-built for just about everything at this point.
As for GPTs or API Assistant, don’t know how you’re going to use them with 50K of documents.
I am working with some 25K of documents in a platform I developed in PHP, so it can definitely be done. But, I came to the game with an extensive development background.
1 Like
now i didn’t decide yet who will develop it just im evaluating the options
i have 50K documents of system and regulations and i want chat with them which option will be good for me ?
Retrieval-Augmented Generation
- Basically, using embeddings to send context documents in question prompts:
- “In Context Learning”
- “RAG” = “Retrieval Augmented Generation”
- What is RAG?
GPT4 Tutorial: How to chat with multiple pdf files - The Chat Completion Process (R.A.G. / Embeddings)
And this is the strategy I used to accomplish the goal, building a platform on top of a content management system designed to maintain thousands of documents: SolrAI Module Development Proposal | Software Creations