The LLM Stack Decision Nobody Makes Cleanly
There is a meeting that happens in almost every team building with LLMs. Someone puts four boxes on a whiteboard: prompt engineering, RAG, fine-tuning, and agents. Then the argument starts. Everyone has read a different blog post. Everyone has a different opinion about what the product actually needs. And by the end of the hour, the team has either picked the wrong thing for the wrong reasons or decided to build all four at once.
I have been in that meeting more times than I can count. This article is what I wish someone had handed me before the first one.
The Problem With How We Talk About This
Most LLM architecture content presents the four approaches as a ladder. You start with prompting, graduate to RAG when prompting is not enough, then move to fine-tuning when RAG falls short, and finally reach agents when everything else fails.
That framing...
Copyright of this story solely belongs to hackernoon.com. To see the full text click HERE
