Writing about AI that has to survive contact with production.
Practical notes on strategy, automation, systems and agents. No invented statistics, no citations we haven’t read.
- AI Strategy6 min read
Why most AI projects never reach production
The failure is almost never the model. It’s scope, data access and the absence of anyone who owns the thing after launch.
Read - AI Automation5 min read
Automate the boring path before you reach for agents
Most work described as needing an agent is a deterministic workflow with three exceptions. Build that first.
Read - AI Strategy5 min read
What an AI readiness assessment should actually produce
A maturity score isn’t a deliverable. A ranked, costed list of workflows is.
Read - AI Systems7 min read
How to buy bespoke AI software without getting burned
What to insist on in the contract, the architecture and the handover, whoever you build with.
Read - AI in Property6 min read
Where AI actually fits in a property business
Enquiry handling, document chasing and compliance paperwork are the realistic starting points. Valuation isn’t.
Read - AI Agents6 min read
AI agents: what has to be true before you run one on live operations
Evaluation, guardrails, observability and rollback. In that order, and before the agent.
Read
Next step
Tell us what the work is costing you.
Five short questions about the business, the problem and the budget. We’ll come back with a view on scope, cost and whether it’s worth doing at all.