AmorphiQ

Quantum reservoir computing is the idea behind AmorphiQ: instead of training a whole quantum circuit, it lets superconducting qubits transform incoming time-series data through their natural dynamics and trains only the classical readout layer. The Cambridge startup, incorporated in July 2025, sells this as a hybrid quantum AI service, pitching it at material degradation modelling and market forecasting. Innovate UK gave it £49,183 for a proof-of-concept running from November 2025 to April 2026, which named trading, fraud detection, grid optimisation and factory anomaly detection as first targets.