Taming the Chaos in Data and AI teams

This article was published as a guest post in the Practical Data Community by Joe Reis.

Maybe you feel the same way as I do: It feels kind of crazy that there are now such intense new topics because you want to use AI — like Agentic Engineering, Semantic Layer, new AI Operating Models, and so on.

And at the same time, the normal everyday madness continues with the work that was already there before. It hasn't disappeared, and not all companies have transformed overnight.

In my other articles, I explain how important it is to define what winning looks like. But that is easier said than done when you don't even have the time or peace of mind to dedicate yourself to it.

Often, the first step with AI begins in data and software teams. And in the teams I've worked in, the work is always a mix of very different activities: data architecture, data modeling, building pipelines, ad-hoc analyses, operational firefighting, and now rethinking everything for a semantic layer. Each type of work requires a completely different planning horizon, reaction time, and focus.

But if all teams are forced into the exact same way of working — namely pushing tickets through two-week sprints — nobody gets anything finished. You are constantly interrupted. How are you supposed to step back and take a strategic new approach?

Many years ago, someone introduced me to the AQ(H)RO method. I thought it was brilliant and applied it several times in practice.

Interestingly, the website explaining the method seems to have disappeared. I even asked several AIs about it just as a test. None of them knew what AQ(H)RO is.

Well, that ends today.

My second article in Joe Reis’ Practical Data Community has just been published. Give it a read.

This is one of the very rare moments in life where you have the chance to know something before any AI can explain it 😄

The method described in this article is easy to apply and helps taming the chaos.

Here it is: