I keep harping on this, but the question is not "can you use your filesystem as a graph database" - of course you can - but whether this performs better or worse than a vector database approach, especially at scale.
The premise of Atomic, the knowledge base project I'm currently working on, is that there is still significant value in vectors, even in an agentic context. https://github.com/kenforthewin/atomic
Cool project.
thanks for checking it out!
Oh neat! Whenever I was working on RAG proof of concepts vector databases seemed to generate noisiest outputs that happened to include my information from my chunks but it was unable to draw reasonable contextual associations. I swap RAG out with a web search tool, all of a sudden the quality goes way up. Is RAG ever going to be easier to hold or should lay people like me just stay moving on?
I think agentic RAG still has its place. a hybrid semantic/keyword search tool in addition to other research tools outperforms the baseline in my experience.