Writing

Notes on getting data science into production

What we have learned building pipelines, models and AI assistants for businesses that need them to work every day. Written for the people who have to sign off on the budget, not just the ones who write the code.

Machine learning

What a churn model actually costs — and when it pays for itself

The arithmetic to run before you commission anything, where the budget really goes, and how to benchmark a model against the account manager's instinct it is meant to replace.

16 August 2026 · 9 min read

Have a problem that looks like one of these?

Tell us what data you have and the decision you are trying to improve. We will reply with whether we can help and what the first step would cost.

Book a call What we build