Is this wholly self-hostable? I'd be curious to run something like this on a home server, have some small model via ollama slowly chew through my documents / conversations / receipts / .... and provide a chat-like search engine over the whole mess.
I have a cron job that executes every night and iterates through my filesystem looking for changes since the last time it ran. If it finds new files or changes, it creates embeddings and stores them in Milvus.
The chat with LLM using Embeddings if not that great yet. To be fair,I have not yet tried to implement the GraphRAG or Claude's contexual RAG approaches. I have a lot of code in different programming languages, text documents, bills pdf, images. Not sure if one RAG can handle it all.
I am using AWS Bedrock APIs for LLama and Claude and locally hosted Milvus