I’ve tried llamaindex a bit because I’m curious if their different methods of indexing improve retrieval. It seems to me the value add llamaindex has is in its unique ways of indexing and chunking. I am new to this, but it seems having to figure out chunking is pretty low level and something that would be better done by “expert software” that figured out the best chunking based on the document (and perhaps rechunk based on the question?). llamaindex seems to do this for me (not fancy probably, but I don’t have to spend time figuring out by chunking strategy). Has anyone had postive or negative results combining llamaindex with langchain? Thank you.
1 Like
I also a bit confused. When I came over this course on LangChain, I noticed that langchain.indexes does basically the same(?) thing as LlamaIndex. Does langchain covers functionality llamaindex? Are there any features in LlamaIndex which would make use both of them?