Course with Felicity
Felicity — RAG vs Fine Tuning: Make the Call
RAG vs fine tuning, decided 1:1 with an AI tutor that takes the opposite side and demands your numbers. Four sessions pricing context engineering, retrieval and a LoRA fine-tune against your own traffic, ending in a one-page recommendation with cost per thousand requests.
What you’ll master — Outcomes you can use
- I present a one-page recommendation with cost per thousand requests and survive ten minutes of hostile questioning on it
Your tutor
A machine learning engineer who has shipped fine-tunes that should have been prompts and retrieval nobody needed.
Analytical · Contrarian · Fond of arithmetic
The plan — What’s inside
- Naming the Failure Type
- Costing the Three Options
- Prototyping the Cheapest First
- Defending the Decision
4 parts
Questions about this course
Should I fine-tune a model or improve my prompt and retrieval instead?
RAG vs fine tuning, decided 1:1 with an AI tutor that takes the opposite side and demands your numbers. Four sessions pricing context engineering, retrieval and a LoRA fine-tune against your own traffic, ending in a one-page recommendation with cost per thousand requests.