Fine-tuning and adaptation
Adapts existing models to a specific domain, and knows when not to.
Fine-tuning is advertised as a skill inside machine learning engineering rather than as a job. The proxy is reasonable for the training half of the work and poor for the evaluation half.
The published figures
One title stands behind this discipline.
Machine Learning Engineer
as the source names it| Basis | Lower (10th) | Middle | Upper (90th) | Sample | Weight |
|---|---|---|---|---|---|
| Contract, per day | £500 | £575 | £738 | 63 rates quoted | thin |
| Permanent, per year | £48,750 | £76,000 | £111,250 | 36 salaries quoted | thin |
A small sample on both sides. Thirty-six quoted salaries is not enough to read a percentile confidently, and the permanent median fell year on year on that thin sample rather than because the job got cheaper.
Outside London the permanent middle is £65,000, against £76,000 UK-wide.
As a fractional engagement
Built on £575 a day, the lowest advertised middle of the titles above. Lowest, not highest, because the number that matters to a buyer setting a budget is the one they are least likely to be surprised by.
| Commitment | Days a month | Per month | Per year |
|---|---|---|---|
| One day a week | 4 | £2,300 | £27,600 |
| Two days a week | 8 | £4,600 | £55,200 |
| Three days a week | 12 | £6,900 | £82,800 |
£575 a day multiplied by the commitment, and nothing else. No seniority uplift, because any uplift would be invented.
Why this one is hard to judge
Knowing when fine-tuning is the wrong answer is worth more than knowing how to do it, and it is harder to demonstrate.
That is the reason a rate alone tells you very little here. Two people quoting the same day rate for this work can be worth wildly different amounts to you, and the difference does not show up in either of their profiles.
Figures from IT Jobs Watch, UK contract rates and IT Jobs Watch, UK permanent salaries, six months to 8 August 2026, read on 2026-08-08. They are rates and salaries advertised, not paid.