We Build the Intelligence Layer for Commodity Decisions
Datasphere Analytics turns global events into defensible procurement and risk decisions,
for the teams that can't afford to be wrong.
Theory and Practice. In the Same Room
Lukas Haemisch | Co-Founder & CEO
Hey, I'm Lukas. At Datasphere, I build the bridge between model and market — making sure what our ML team builds actually lands with the people who need it most: procurement managers and finance teams who have decisions to make every day.
My job is to translate the intelligence layer into products that feel intuitive, not technical. I'm constantly asking: what does a Head of Procurement actually need to see at 8am on a Monday? What makes a forecast defensible in a CFO meeting? What separates a tool people open once from one that becomes part of how a team thinks?
I keep our organization focused on what customers need to decide — not on what technology can theoretically do. Because in the end, the best model in the world is worthless if the person who needs it doesn't trust it, doesn't understand it, or doesn't have time for it.
Patrick Petter | Co-Founder & CPO
"Hi, I'm Patrick. I lead product strategy and machine learning at Datasphere — which in practice means I sit at the intersection of two very different worlds: the world of models, pipelines, and training data on one side, and the world of procurement cycles, hedging decisions, and CFO dashboards on the other.
My job is to make sure those two worlds actually talk to each other. A model that achieves 99% accuracy in a notebook means nothing if the output doesn't arrive at the right moment, in the right format, with the right context for someone to act on it. I obsess over that last mile — the point where intelligence becomes action.
What drives me is a simple belief: a model is only as good as the decision it enables. That's the standard I hold our ML work to, and it's what gets me out of bed in the morning.
Prof. Dr Reiner Kurzhals | Co-Founder & CRO
Hi, I'm Reiner. My background is in quantitative methods and applied research — and my work has always revolved around one question: why do organizations struggle to make good decisions, even when they have access to good data?
At Datasphere, I shape the forecasting models, the architecture, and the intellectual framework behind how we think about AI in industrial decision-making. Accuracy is never the end goal — it's the baseline. What matters is whether a forecast actually changes a decision, and whether the person making it can stand behind it.
My deep conviction: trust in AI is built through transparency, not through black-box accuracy claims. That belief runs through every model we deploy, every confidence range we show, and every assumption we make explicit. AI that can't explain itself has no place in a procurement committee."
Three Principles Behind Everything We Build
"Accuracy is hygiene. Defensibility is the buying reason."
A forecast that's right 99% of the time is only valuable if the decision-maker can stand behind it. We build for accountability, not just precision.
"Volatility doesn't drain capacity. Noise does."
The bottleneck in commodity decisions isn't market complexity — it's the hours spent collecting, sorting, and re-explaining information that should already be structured.
"Safe change over better tech."
Organizations don't adopt intelligence tools because they're powerful. They adopt them because they feel safe. We design for low-regret entry — into routines that scale, not disruptions that stall.
Built for a Problem We Kept Seeing
We didn't start Datasphere because AI was trending. We started it because procurement and risk teams deserved better than morning routines built around five browser tabs and a gut feeling. ~ Lukas Haemisch | Co-Founder & CEO
The problem was never a lack of data. It was a lack of signal — structured, causal, actionable intelligence that arrives before the decision, not after the damage. That's what we set out to build.
~ Patrick Petter | Co-Founder & CPO
Industrial companies spend billions on raw materials. Most of them make those decisions with incomplete information, supplier-sidedata and instincts sharpened by experience — but not by intelligence.
~ Prof. Dr Reiner Kurzhals | Co-Founder & CRO
We Pubslish
What We've Learned
Why Good AI Still Fails
AI doesn't fail because the model is bad. It fails because it disrupts existing operating systems — routines, tools, roles. The right framing: AI doesn't compete with experts. It frees them from functioning as human search engines.
Why AI Makes Decisions Harder
AI cheapens analysis. It makes accountability expensive. When answers are instant, the old ritual of slow, expensive analysis as decision-cover collapses. What remains: ownership. Accountability. Decision.
Stop Preparing. Start Deciding.
The bottleneck was never information — it was trust. AI doesn't replace the need for trust. It forces organizations to build it faster and more explicitly.
From Better Tech to Safe Change
Organizations don't buy intelligence. They buy safety. "Better tech" is the wrong pitch. The successful pitch is: we're safer. Try this.
Operating from
Around thr World
A conversation to begin - Let's explore what's possible
No obligations, just a conversation.
