[D] Real talk about RAG

Let’s be honest here. I know we all have to deal with these managers/directors/CXOs that come up with amazing idea to talk with the company data and documents.

But… has anyone actually done something truly useful? If so, how was its usefulness measured?

I have a feeling that we are being fooled by some very elaborate bs as the LLM can always generate something that sounds sensible in a way. But is it useful?

data deal documents machinelearning managers rag something talk the company