Most AI projects fail not because of the technology — but because the solution doesn't fit the organisation. Anyone can try an AI model on the internet today. But AI that actually helps has to work with your data, on your systems, and for your processes.
The AI has to adapt. Not you.
Entropy is me — Dr. Bastian Entrup, a certified IBM Level-2 Architect. I hold a doctorate in computational linguistics with a focus on machine learning, and then spent over ten years at an internationally operating software and consulting group — as lead architect and project manager for large banks, insurance companies, energy companies, and European supervisory authorities.
In that time I designed and delivered systems that are now running in regulated environments: LLM-powered document analysis, knowledge-graph integration into existing database landscapes, large-scale NLP pipelines for customer communications, and data-lakehouse architectures spanning dozens of data sources. I have led development teams of up to twenty engineers and taken technical programmes to completion over multi-year engagements.
Before moving into industry I taught computational linguistics and machine learning as a research assistant at university, and published in peer-reviewed conference proceedings and journals — including Springer and ACM.
I come from a family with many self-employed people and I can see that this AI revolution hasn't yet reached the Mittelstand — even though it's been entirely feasible for some time. What used to require months of development time and a budget of several hundred thousand euros can now be realised in a matter of weeks. The knowledge I built up in large corporations, I'm now bringing to the Mittelstand.
The name says it all: in information theory, entropy stands for complexity and information density. Your data contains more knowledge than you realise — and I want to help you unlock it.
