Blog
Notes from production
Writing from the team on AI agents, data platforms, and what it actually takes to run AI systems in production. Long-form, with the working shown.
Data Hatch
Data Hatch
Writing from the team on AI agents, data platforms, and what it actually takes to run AI systems in production. Long-form, with the working shown.
The Act reaches some UK insurers, not all, and the high-risk deadline is now 2 December 2027. Who is in scope, what the Digital Omnibus changed, which insurance AI is likely high-risk, and a checklist to work through.
Read the guide ↗Processing, handling and adjudication are three different jobs, and most claims automation only does the first. What straight-through processing really means, what UK rules allow, and how to judge AI claims management software.
Read the explainer ↗One doctor in a hundred carries a third of the claims, yet every application waits in the same queue. How a triage engine routes each case, modelled on a book of 20,000 doctors.
Read the analysis ↗Why AI bills rise while model prices fall, and how four controls plus a shared data foundation hold a two-year cost envelope. A modelled analysis.
Read the analysis ↗