Questions about AI in your organisation

Four questions that separate a useful AI partner from a sales pitch, and what the right answers sound like.

Four types of AI provider in the Dutch market, what each one is good for, and how to tell which one you actually need.

Four things that separate AI implementation built for SMEs from a programme designed for enterprises.

You do not need a data scientist to start with AI. You need three other things, and the technology comes last.

One process, four weeks. A concrete way to start with AI adoption that tells you within a month whether it works.

Three levels of AI investment, from a 25 euro subscription to custom automation, and which one you should actually start at.

The three customer service tasks where AI actually pays off, and why a chatbot on your mailbox is almost never the answer.

Where AI pays off in marketing and back office work: preparing the work, with a human keeping final control.

Separate the tool from the data and the safety question becomes manageable. Three categories, one working session, one page of rules.

People adopt a tool that removes an annoying job. Start with what costs them time, not with what AI can do.

Three reasons AI projects stall: started too big, no KPIs agreed up front, and the knowledge stayed with the supplier.

In technical services the AI gains sit in looking things up and in the admin, not in the work on site.

AI is reliable for work you check and unreliable for work you send out blind. What that means in practice, and what Air Canada learned the hard way.

Building a custom GPT takes half an hour. Knowing when it is the wrong answer saves you a great deal more.