Quantum lab

The pricing decision is written as a QUBO — one binary variable per product-and-price pair — and solved with quantum-inspired annealing. Every run is benchmarked against an exact classical solve so you can trust the answer.

Optimality gap
0.0000%
quantum vs exact objective
Identical price picks
19/19
Annealing time
0 ms
Exact solve time
0 ms

Problem formulation

Binary price-selection variables with hard and soft constraints.

Variables. x[p,k] = 1 when product p is sold at candidate price k (five candidates per product, 0.90× to 1.10× of today's price).

One price per product. Enforced as a one-hot penalty A·(Σ x[p,k] − 1)² and preserved structurally by the annealer's move set.

Minimum margin. Candidates below the margin floor are removed before the QUBO is built, so infeasible prices cannot be selected.

Price stability. A quadratic penalty pulls the average price index toward 1.00 and a band penalty keeps it inside 0.98–1.02.

Objective. Maximise weighted profit, inventory health and customer value; subtract waste and price movement.

Annealing energy per restart

Lower energy is a better solution.

Benchmark across runs

Gap in percent against exact solve time.

Bubble size is annealing time. A gap of 0% means the quantum-inspired solver reproduced the exact optimum.