AI workflows
Where should an AI workflow stop and ask a person?
Design each approval around what the action changes. Give the reviewer enough context to make a real decision.
Read the articleInsights
Notes from AI engineering
Practical notes on workflows, company knowledge and building AI systems people can run.
Questions worth asking before you build
Knowledge AI
An answer is more useful when you can see where it came from
Start knowledge AI with real questions and authorised sources. Then give people an easy path back to the evidence.
Read the articleDelivery and enablement
Design the handover before the first release
The people who run a system need to know why it was built that way and how to recover it. Make that part of the build.
Read the articleYour next step
Choose one workflow to improve
Tell us which workflow you want to improve and where it gets stuck. We’ll talk about whether a design sprint is the right place to start. Then we agree the scope, deliverables and price before any work begins.