Dr. Gabriel Lipnik
AI & Optimisation Engineer
Railway, logistics & enterprise AI
AI & Optimisation Engineer
Railway, logistics & enterprise AI
If you're running complex logistics and wondering whether AI can actually help, let's find out. The interesting problems are the ones where the answer isn't obvious.
I build AI systems that make logistics and supply chain operations measurably smarter. Where wrong decisions are expensive and right ones are hard to compute.
My background combines a doctorate in technical mathematics from TU Graz (awarded sub auspiciis Praesidentis) with training in AI strategy (MIT Sloan) and AI governance (Oxford Saïd Business School). I work at the intersection of mathematical optimisation, agentic AI, and production engineering, where the goal is not a working prototype, but a system that actually changes how decisions get made.
At Anexia, I focus on transport and logistics domains where planning decisions have real operational consequences. I teach mathematics at the University of Klagenfurt, bridging theory and practice.
AI-driven decision systems and optimisation · Explainable and trustworthy AI · AI governance and responsible deployment · ML and GenAI in operational contexts
Discrete mathematics and combinatorics · Applied optimisation (mixed-integer, linear) · Algorithmic thinking and formal modelling
Saïd Business School, University of Oxford
2025
AI governance, EU AI Act, regulatory frameworks, and responsible AI.
MIT Sloan School of Management
2025
AI strategy, organisational implementation, and business impact.
Federal President Alexander van der Bellen
2024
Highest academic honour in Austria.
Austrian Federal Ministry
2020
National award for outstanding master's thesis.
I'll be speaking at PyData London 2026 about what the EU AI Act actually means for the day-to-day work of ML engineers. A full recap with slides and takeaways will follow after the event.
Read more
AI coding agents are rapidly becoming part of the developer stack. But while the ecosystem moves fast, one question is becoming critical in regulated environments: who actually controls what these agents do at runtime?
Read more
One of the most impactful talks I attended at Codemotion Milano 2025 showed how semantic routing, vector similarity search and semantic caching can dramatically reduce token usage, latency and energy consumption, while keeping quality and context intact.
Read more