Research behind the platform
Aipokit began as a method, not a product. Its founder, Dr Jürgen Wöckl, holds a PhD in Physics and Mathematics from TU Wien, spent time at McKinsey, and has lectured for fourteen years at WU Vienna, FH, and as a ten-year adjunct in Bangkok. The through-line is agent-based modelling: a rigorous way to reason about systems of interacting agents. The platform is what happens when that method moves from lecture to production system. The research below is not an appendix — it is the lineage of what ships.
Systems are made of interacting agents.
Agent-based modelling simulates a system from the bottom up — many autonomous agents, each with local rules, producing behaviour no single equation predicts. It is the discipline behind Dr Wöckl's published work, and the conceptual lineage of Aipokit's agent runtime. Where the academic model simulated agents to understand a system, the platform runs agents to operate one. The intuition is the same; the target moved from paper to production.
When the right answer beats the fastest one.
Much of the applied work sits in decision support: planning where the answer must be defensible, not merely quick. That means reasoning over constraints, weighing scenarios, and making trade-offs legible rather than hiding them behind a confident guess. The clearest instance is pharma operations planning — where a wrong schedule costs more than a slow one. See how the method reads in practice on the <a href="/en/operations-planning">operations planning story</a>.
From research to production
Physics and Mathematics, TU Wien
A grounding in formal systems and the mathematics of how they behave — the toolkit that makes agent-based reasoning rigorous rather than merely intuitive.
14 years lecturing — WU Vienna, FH, Bangkok
Fourteen years teaching modelling and decision methods, including a decade as a Bangkok adjunct. Teaching is where a method is stress-tested for clarity — the same clarity now carried into training.
Agent-based operations planning in production (pharma)
The method leaves the lecture hall: agent-based planning running against real pharma operations, where the quality of a decision has a measurable cost.
Aipokit — the platform generalises the method
What was a bespoke planning system becomes a general platform — an agent runtime, a memory graph, and a process engine any organisation can run on its own infrastructure.
Applied research that feeds the platform.
The published work on agent-based modelling is not shelved history — it is the source material. The same ideas that appeared in peer-reviewed form now inform how the agent runtime is designed and how the methods are taught. Research feeds the platform; the platform feeds training, where practitioners learn modelling and decision methods first-hand.
Collaborate or cite?
If the research is useful to your work — a collaboration, a citation, or a conversation about method — we would like to hear from you. The platform is applied; the ideas behind it remain open to discussion.