Most of the tasks that we solve are classified as NP-hard. There are no search algorithms that provide an exact solution in reasonable time. Theoretically, exact solutions can be found using quantum computers, but current models aren’t up to scale to solve practical tasks.
That’s why we use the ideas of discrete mathematics and quantum computing to efficiently solve these tasks on currently available classical hardware with 99% accuracy, but in reasonable time.
We’ve developed an algorithm called QuSolve SpaceCut. It helps to preprocess raw source data allowing to reduce the search space and smooth the landscape of the cost function.
QuSolve Heisenberg Machine solver finds the optimal solution. In more complicated cases it is used in combination with ad-hoc equipment.