Glossary
Supervised fine-tuning
Quick answer
Fine-tuning with labeled input and output examples showing the desired response.
SFT is often the first post-training step.
Real expert responses, like resolved support answers, can supply examples.
Why Supervised fine-tuning matters when licensing company data to AI
Supervised fine-tuning, or SFT, trains a model on labeled input and output examples that show the desired response. It is often the first post-training step. Real expert responses, such as resolved support answers, can supply those examples, so buyers check that each response reflects real, completed work.
- 01Your work records
- 02Personal details removed
- 03You approve the release
- 04Licensed to AI developers
Related terms
Related resources
- IndustryBPO & contact centers data
- IndustryFintech software data
- QuestionDo AI labs buy medical data?
- QuestionDo AI labs buy video of people working?
- InsightHow do I de-identify customer support transcripts for AI training?
- InsightHow do I de-identify dispatch logs for AI training?
- InsightHow do I de-identify IT service tickets for AI training?
- SolutionAI training data: what it is and where the best of it comes from
- SolutionWhat is AI evaluation data?
- DataCode review records
- DataCompany email archives
- GlossaryRevenue share
See if your company qualifies
A short company assessment. No data uploads are needed.