Insights
Thinking out loud
On AI engineering, regulated environments, and why the way software gets built inside financial services is about to change.
How regulators evaluate XBRL platforms
The criteria that separate good XBRL platform decisions from expensive ones: data model, taxonomy change cost, validation ownership and the exit question.
Supervisory data collection: 7 questions to ask
A practical checklist for supervision teams evaluating data collection technology, from validation ownership to granular data readiness.
Why your regulated software project is 6 months late
It is not the regulation. It is not the compliance team. It is the delivery model you chose before writing a single line of code.
AI is not replacing your engineers. It is replacing your delivery model.
Every consultancy is talking about AI copilots. Almost none of them are asking the right question: what happens when AI changes how teams are structured, not just how code is written?
What a 2.5x ROI taught us about building for regulators
We built four production systems for a global financial regulator. The results were independently verified. Here is what we learned.