insight
What I learned shipping agentic systems in production
18 April 2026
Notes from eighteen months building agentic AI for a multi-brand operating business — what worked, what didn't, and where the wall actually is.
Most agentic AI write-ups are demos. This isn't. The systems below have been live, in production, inside an operating business with real customers and real revenue, for over a year.
What we shipped
Placeholder section — Richard to write up the specifics. The list includes a content production pipeline, a multi-brand pricing optimisation loop, a sales-research agent, and a tier of internal copilots wired into our ERP.
What worked
Placeholder section.
What didn't
Placeholder section.
The wall
The hard part isn't the agent. The hard part is the data, the observability, and the change management around the agent. If you don't have those three, the agent is a demo, not a system.
More on each of those — and how to build them without a six-month detour — in upcoming posts.
Get Actionable AI in your inbox.
One practical AI play per issue. Sent occasionally, never filler.
Ready to use AI seriously?
A 30-minute call. No deck, no follow-up nurture sequence. I'll tell you whether I can help.