I would like to understand if anyone has implemented generative AI without using openAI and if so which open source you have used and how successful it has been so far
We have enterprise level incident data, relevant documentation etc that users will search about and need to generate responses using generative ai.
Is it possible to do this without relying on open AI at all
If you’re “searing” specific documents and incidents, would it make more sense to use either vector search or a retrieval (RAG) model to pull from these sources, instead of generating new text? Even if you have a lot of data it may be difficult to fine-tune LLaMa to generate reliable responses.