Machine Learning Engineer

We are less interested in state-of-the-art benchmarks than in a model that behaves predictably on a client’s messy real data, and keeps behaving that way six months later.

What you will do

• Design, train and evaluate models against data that is never as clean as the brief said
• Build the evaluation harness before the model, so improvements can be proven
• Deploy to production and instrument enough to notice when quality degrades
• Write up what you tried that did not work, because that is half the value

What we are looking for

• Two or more years of applied ML in production, not only research
• Strong Python and the usual stack
• Experience evaluating model quality with something more rigorous than eyeballing outputs
• Comfortable saying when a problem does not need machine learning

How we work

Small senior team, no middle layer. You talk to the client and to the founder directly, and you own what you ship from scope through to production.

We reply to every application, including the ones we turn down. If we are not hiring for this exact role when you apply, we will say so rather than leave you waiting.