Building & tuning
Powerful, but usually the wrong first move. For most business problems, retrieval and prompting are cheaper and far easier to update than a fine-tuned model.
In plain terms
Fine-tuning is further training a foundation model on your own examples so it adopts a specific tone, format, or task.
Fine-tuning means taking a model that already knows language and giving it extra practice on your examples, so a specific habit becomes automatic. People use it for tone, for a fixed output shape, or for a judgment that prompting does not hold steady.
It is a later step, not the first one. If the answers already live in your documents, show the model those documents for each question. If a careful instruction already works on your test cases, stop there. Fine-tuning costs money, needs clean examples, and has to be redone when the job changes. It also does not add facts the examples never contained.
Think of it this way: Fine-tuning is taking a highly educated generalist and putting them through a specialist apprenticeship using your own documents, tone, and processes.
A customer support team fine-tunes a model on 10,000 of their own resolved tickets. The result handles tier-1 queries with the company's exact phrasing and escalation logic, without any prompt engineering overhead.
A support team wants every summary to start with the customer's ask, then the promised next step, then the order number. Instructions get them close, but the format drifts. A fine-tune on three hundred corrected summaries makes the shape stick. The facts still come from the ticket, not from the tune.
When a specific tone, format, or domain vocabulary is critical and cannot be achieved through prompting alone. Works best when you have hundreds of high-quality examples of the target output. Not as a first step. Start with clearer instructions, and with answers pulled from your own documents. Fine-tuning costs more to set up, needs strong examples, and is slow to change when your policies change. Higher setup and upkeep cost than clearer instructions or searching your own documents.
RaftLabs points the model at your documents and your rules, then checks the answers against cases you already trust. That surrounding work is where these projects succeed or stall. The related work on our side is LLM fine-tuning.
This sits with the other building & tuning terms on the glossary. How a general model gets pointed at your documents, your tone, and your workflow. Worth reading next: Retrieval-Augmented Generation (RAG), Prompt Engineering, and Embeddings.
Work with us
Tell us what's not working in your business. We'll find the real problem and tell you exactly what it would take to fix it.