Explore what a decision does against a virtual clock, across a fleet, before it reaches anything real. Simulation is the deterministic form of finding out in production — the same information, arriving early enough to be worth having.
A digital twin mirrors the business; what comes back is foresight. Every result replays exactly from four values: engine version, scenario, seed, replication index.
The consequence of a change is learned from the change. The evidence arrives after the cost, it is contaminated by everything else that happened that week, and running the experiment again is not an option.
The same question is answered against a mirror of the business, across a fleet, at whatever speed the clock is set to. Rules are declared as data, checked before a run and watched on every step, and any result replays exactly.
Engine version, scenario, seed, replication index. Four values reproduce a result — which is what makes a simulated finding something you can put in front of someone who doubts it.
The constraints of the modelled world are declared, not coded into the simulation loop. They are checked before a run starts and watched on every step, so a twin cannot drift silently away from the thing it mirrors.
The engine underneath Polytrion has been qualified against nine published answers in nine fields where the exact answer was already known — not against another simulator's output.
The same world that agrees with the reference tools has also been asked something nobody had the answer to. That produced a preprint on backlog metastability in windowed quantum-error-correction decoding.
A digital-twin design platform: rules declared as data, checked before a run and watched on every step, with results that replay exactly. Ships a Monte Carlo simulator, with the engine certifying the catalogue it runs on. Soft launch, by invitation.
Polytrion → The QEC paper what came out of a twin“Backlog Metastability in Windowed Quantum Error Correction Decoding” — research that came directly out of modelling a quantum computer in the engine, from world construction through experiments to authoring and review.
Read it on Preprints.org ↗