Glad you've enjoyed the documentation so far! More education is definitely the plan.
Many people are new to the LLM space, so spending time explaining the theory and mechanisms behind LlamaIndex abstractions not only helps people understand how they work, but also shows how much time LlamaIndex is saving you from writing stuff from scratch 💪
Absolutely. And seriously, LlamaIndex is in one of /the/ use cases for LLMs. I really believe RAG, whatever the store underneath, is one of the killer apps. Everyone wants it. The company I work for set me to investigating options internally for tooling and they're starting to get clients asking for solutions incorporating it.
But the demand/need, either as part of a bigger solution or essentially the product itself, is just enormous. It's one of the more exciting areas of LLMs to me, that and agents/multi-agent systems. Which will need LlamaIndex to be grounded.
Another thing though? Getting through to VP level engineering decision-makers. I've been running into headwinds where people want to build their RAG pipelines from scratch because they're worried the abstractions aren't stable and the players today will be gone tomorrow.
