What to ask an AI vendor before signing anything
The questions to put to any AI supplier about data, cost, control and exit before you sign, and the answers that should worry you.
Insights
Notes from the work.
Plain writing on agentic AI, commerce and research: what works, what does not, and what it costs.
Evaluating an agent: tests that should pass before real work
A practical set of tests every AI agent should pass before it touches customers, money or records, and how to build them from your own cases.
The morning report nobody has to write
How an agent can gather yesterday's figures from your systems and deliver a short, reliable morning report before the day starts.
Cash on delivery and agents: confirming orders before the courier leaves
How an AI agent can confirm cash on delivery orders before dispatch, cut returned parcels, and hand the unclear cases to a person.
Tools are what make an agent useful
Why the tools an agent can use matter more than the model behind it, and how to design, name and limit them well.
Build, buy or wait: an honest decision guide for AI projects
How to decide whether to build a custom AI system, buy an existing tool, or simply wait, with clear signs for each choice.
Negotiating agents: an old research field with new relevance
Researchers studied software that bargains long before chat models. As agents start buying and selling, their ideas matter again.
A knowledge assistant for staff: answers from your own manuals
How to give staff an assistant that answers from your own manuals and procedures, cites its source, and admits when it does not know.
WhatsApp orders without the chaos: what an agent can take over
Which parts of WhatsApp ordering an AI agent can handle reliably, which parts need a person, and how to set it up without losing the personal…
How an AI agent decides what to do next
A plain walk through the loop inside an AI agent: what it looks at, how it picks a step, and where its choices go wrong.
How to pick the first process to hand to an agent
A simple scoring method for choosing the first job to give an AI agent, so your first project is small, safe and easy to judge.
Coordination: how several agents avoid tripping over each other
When several AI agents share one job, they duplicate work, clash and loop. Here are the coordination patterns that prevent it.