Quantum-accelerated risk estimation — a proof of concept, without the hype
A working proof of concept that recovers a real economic tail probability with quantum amplitude estimation on Comuvia's simulator — paired with an honest verdict of forward-readiness, not a production speedup today. The point is being able to tell hype from reality on your own stack.
Outcome metrics
- QAE recovers the Minsky tail probability
- 0.4703 → 0.4700 (matches the classical Monte Carlo result)
- Backends
- NumPy statevector (CI) + CUDA-Q / cuStateVec GPU hook
- Honest verdict
- research / forward-readiness — no production speedup yet
- Method
- quantum amplitude estimation over a real shock-MC tail
What it is
A working quantum-amplitude-estimation (QAE) prototype that re-derives a real economic tail probability — the recommendation-flip from Comuvia's Minsky-family shock simulation — and checks it against the classical Monte Carlo answer.
The result
QAE recovers the tail probability to 0.4700 vs the classical 0.4703 — it reproduces the real number, validating the quantum encoding of an actual decision-relevant quantity. It runs on a NumPy statevector in CI and has a CUDA-Q / cuStateVec hook for the on-prem GPU.
The honest verdict
No production speedup today. On current simulators and near-term hardware this is forward-readiness research, not a faster risk engine. Saying so plainly is the deliverable: the value is a vendor-neutral team that can evaluate frontier technology on its own stack and separate genuine capability from vendor hype — before a buyer spends on it.
Why it matters for a buyer
For a leader weighing frontier-tech bets, Comuvia offers practitioner-grounded scenario analysis: we build the PoC, recover a real number, and tell you whether the timeline is "now," "watch," or "hype" — with the working code behind the call.