Leadership Team

Our story

From physics to real-world impact

Andre and Peter met in 1998 as doctoral students at a Max Planck Institute near Potsdam. Both earned PhDs in theoretical physics (soft condensed matter) and were trained to deliver rigorous analysis that holds up under scrutiny.

Their professional paths then split.

Peter stayed in academia and became a professor in Marburg, Germany. Andre moved into industry, first in a large company (BASF) and later into a startup (Novihum). Despite the distance, the friendship held.

The call that started it

In 2017 Andre called Peter with an urgent problem: the agricultural trial data from Novihum was too noisy and inconsistent to support a key product claim. If the team could not prove the effect, there would be no credible story, and definitely no chance of winning customer trust and developing its market.

So they did what physicists do: questioned assumptions, dug into how the data was generated, and rebuilt the analysis from first principles. The result was a cleaner view of the data, and many insights that had been hidden in plain sight.

Despite proving the product’s extraordinary properties, that startup did not work out. But a lesson stuck: a defensible, clean analysis of product effect is the minimal condition for any customer conversation. It is what gives a sales team confidence in the product. And many teams struggle with exactly that step: turning messy real-world data into evidence they can stand behind.

The bigger pattern

We started lyfX.ai in early 2024 and from the beginning worked with teams across agriculture, chemicals, and manufacturing. The problems we saw were always very similar: messy data and fragile analysis that affected really important business decisions: claims, regulatory questions, budgets, reputation.

The teams were not short on data. They were short on time, tools, and a repeatable way to turn that data into conclusions they can defend.

From consulting to building

At first, we worked hands-on as data specialists alongside a team, understanding the process, cleaning the data, building the analysis, and delivering results they could trust. We still do this kind of consulting work, but a pattern emerged and the focus shifted.

We were solving structurally similar problems again and again. The domains changed, the data changed, but the core task was the same. That is when the focus shifted. Instead of doing the analysis for each client from scratch, could we encode what we know into tools that teams use themselves?

This is how we have always worked, and it is the same instinct that drives a startup. You run into a problem that blocks real work. You solve it because you have to. And then you notice you are not the only one with that problem. The tool comes after the solution, not before it.

Where we are now

We are a small team, and deliberately so. We combine deep industry experience with modern AI engineering to build tools for process industries. Not generic AI products that need months of customization, but focused solutions for people who work with real processes and real data.