Abel is building a world-dynamics foundation model that learns how systems change, what drives them, and what happens when we act.
The next foundation model will not only understand what the world says. It will understand how the world changes.
thousands of variables, coupled across time, responding to every intervention.
It helps uncover the drivers and relationships inside complex systems — providing structural evidence for how models understand what matters.
This is the next foundation model we build.
Every market, grid, production line, and cell is still modeled separately — a team rebuilds the whole stack for each one, then rebuilds it again when the regime shifts.
Abel is built to identify a system from its measurements and context instead.
Measurements and context from every source become one picture of where the system is now.
RSI Loop — Feedback loop
Our current work spans models, systems and learning loops that move us toward this foundation model.
Find which parts of a system drive the others, and see what changes when one of them moves.
Explore AndreyLearn how systems evolve across time, scale and regime, then roll them forward into possible futures.
Coming soonWe look for people who have gone unusually deep somewhere: a medal, a theorem, a benchmark, a system that survived production.
We work with domain partners in markets, energy, industry, biology and science — especially where systems are wide, decisions are consequential, and real outcomes can close the learning loop.