I fine-tuned an LLM for a client, then told them not to use it

description: "A real client case study on supervised fine-tuning vs RAG and few-shot prompting, with...

giovedì 4 giugno 2026 New tab

1,231 words~6 min read

description: "A real client case study on supervised fine-tuning vs RAG and few-shot prompting, with the actual Azure bill that decided it."

tags: ai, machinelearning, azure, llm

A client asked me to fine-tune an AI model for them. I built it, evaluated it properly, and then recommended they not deploy it. This is the write-up of how I got to that decision, including the real Azure bill that settled it.

The full repo, with the cost breakdown, the before and after outputs, the prompts, and a redacted dataset sample, is here:

https://github.com/xaphor/landscaping-llm-brandvoice-eval

I fine-tuned an LLM for a client, then told them not to use it

I fine-tuned an LLM for a client, then told them not to use it

Other newsrooms on this story

Related reading

LLM Fine-Tuning vs RAG: A Production Decision Framework for Engineering Teams

Fine-Tuning LLMs for Multi-Turn Conversations: A Technical Deep Dive

Building a personalized code assistant with open-source LLMs using RAG…

RAG vs Fine-Tuning- Choosing Right Strategy for Modern AI Applications

Fine-Tuning is Dead: Why Context Orchestration Won in 2026 | Towards AI

How I Cut My AI Bill by Caching LLM Responses in Node.js

Other newsrooms on this story

Related reading

LLM Fine-Tuning vs RAG: A Production Decision Framework for Engineering Teams

Fine-Tuning LLMs for Multi-Turn Conversations: A Technical Deep Dive

Building a personalized code assistant with open-source LLMs using RAG…

RAG vs Fine-Tuning- Choosing Right Strategy for Modern AI Applications

Fine-Tuning is Dead: Why Context Orchestration Won in 2026 | Towards AI

How I Cut My AI Bill by Caching LLM Responses in Node.js