Tokens are cheap. LLMs are fast. Pre-processing and post processing are the real bottlenecks. I know you are going to say that why not Use LLMs for that. Comple

Tokens are cheap. LLMs are fast. Pre-processing and post processing are the real bottlenecks. I know you are going to say that why not Use LLMs for that. Complexity in an end-to-end workflow is a zero-sum game. If you throw more of that workflow to LLM, more complexity comes back to you, to those steps that you need to do on your own. If you keep only 10% of work for yourself, it's going to be 10 times more complex and rapid than what you usually do.