Pilot
Regional LTL carrier · ~1,200 vehicles — 6.2% lower routing cost across 3,418 stops, every time-window and capacity constraint honored.
Now piloting · QuGradient Quantum Optimizer
Quantum optimization for complex decisions.
The hardest decisions in your business are complex for one reason: the problems beneath them are NP-hard. The QuGradient Quantum Optimizer solves them — six solver engines, eight industries, running on the infrastructure you already own.
Spun out of QuMatrix · backed by 350+ peer-reviewed papers and a decade of applied optimization.
Routing cost vs current solver−6.2%
Stops planned3,418
Constraints honored100%
See it work
A schematic walkthrough of the Optimizer.
Problem in, engine selected, solution benchmarked — in ninety seconds.
The science behind QuGradient — research & engagement network at QuMatrix
The opportunity
Operational complexity is outgrowing traditional optimization.
Routing a fleet, pairing a crew, dispatching a grid, rostering a hospital, constructing a portfolio, allocating spectrum — different vocabularies, the same NP-hard mathematics. Classical solvers return "good enough, eventually," and at enterprise scale the gap between good enough and optimal is worth tens of millions a year. QuGradient closes that gap with quantum optimization that holds up under your real constraints.
6%+
Routing cost reduction with QG-ROUTE
preliminary · vs current solver, on partner data
13.1%
Scheduling improvement in benchmark
preliminary · documented methodology
6
Solver engines on one core
15yr
Of optimization research behind it
One Optimizer
One Optimizer. Six engines.
QG-ROUTE
In what order, and along which path? Routing, vehicle routing, pickup-and-delivery, sequencing.
Beta6%+ · preliminary
QG-SCHED
Who does what, and when? Job-shop scheduling, crew pairing, rostering, DAG workflows.
Beta13.1% · preliminary
QG-ALLOC
Where does everything go, and how much? Bin packing, facility location, dispatch, capacity planning.
Preview
QG-SELECT
Which subset do we pick? Portfolio construction, knapsack, feature selection, candidate screening.
Preview
QG-CLUSTER
What are the natural groups? Max-cut, graph partitioning, segmentation, anomaly isolation.
Design partner
QG-GRAPH
How is the network built, and who gets what? Graph colouring, spectrum assignment, topology design.
Research-stage
Why it's different
Four things, working together.
One optimizer, every industry
Six engines built against problem archetypes, not verticals. An airline pairing crews and a hospital rostering nurses are the same NP-hard problem in two vocabularies.
Quantum-inspired & quantum-hybrid
QUBO/Ising formulations, quantum annealing, variational circuits, quantum walks, and quantum-inspired evolutionary and swarm search.
Runs on what you already own
HPC/SMP, GPU quantum simulators, quantum annealers, gate-based QPUs. Your execution target is a runtime setting, not an architectural commitment.
You bring the problem
You supply data, objective and constraints in your own terms; QG-CORE classifies, formulates, selects the engine, and returns a solution benchmarked against your baseline.
Where it's applied
Eight industries, one Optimizer.
Transportation & Logistics
QG-ROUTE · QG-SCHED · QG-ALLOC
Supply Chain & Manufacturing
QG-ROUTE · QG-SCHED · QG-ALLOC
Telecom & Computer Networks
QG-GRAPH · QG-ROUTE · QG-SCHED
Finance
QG-SELECT · QG-CLUSTER · QG-ALLOC
Energy & Utilities
QG-SCHED · QG-ALLOC · QG-GRAPH
Healthcare
QG-SCHED · QG-ALLOC · QG-ROUTE
AI & Foundational Models
QG-SELECT · QG-SCHED · QG-CLUSTER
Bioinformatics & Medicine
QG-CLUSTER · QG-ROUTE · QG-SELECT
How it works
We don't ask you to take the number on faith.
STEP 01
Submit
You provide the business problem, your data, objectives and constraints.
STEP 02
Classify
QG-CORE matches the problem against the archetype library.
STEP 03
Formulate
It's compiled into canonical QUBO/Ising form.
STEP 04
Select & solve
The right engine, algorithm variant and backend are chosen and executed.
STEP 05
Return
You receive a solution, benchmarked against your own baseline.
The platform
The orchestration core under every engine.
One orchestration engine powers all six solvers. Hybrid quantum-classical and quantum-inspired solvers, where they measurably win; classical methods where they don't. We don't claim a quantum computer is in the loop when it isn't.
ExperiencePlanner UI, dashboards, configuration
Optimization servicesQG-ROUTE · QG-SCHED · QG-ALLOC · QG-SELECT · QG-CLUSTER · QG-GRAPH
QG-COREOrchestration — classification, formulation, engine selection, backend routing
Data & integrationConnectors to your operational and data systems
FoundationSimulator · benchmarking infrastructure · the QuMatrix research base
Why now
Quantum-optimized results, ready to deploy.
QuGradient is the deployable form of QuMatrix's research: the algorithms are mature, the benchmarks are documented, and quantum optimization is proving its value in pilots today — ready to run against your real constraints, in any of eight industries.
Proven science
Built on QuMatrix — 350+ papers and a decade of applied optimization across energy, logistics and beyond.
The research →Measurable proof
QG-ROUTE and QG-SCHED are the most mature engines; every claim is tied to a benchmark on real data, labeled honestly.
QG-ROUTE →Ready to deploy
Bring a slice of your data and we benchmark the Optimizer against your current solver — then connect what wins to your operations.
See the research →Two ways to start
See it on your data.
Bring a slice of your data and we'll show you the delta in a 30-minute call — the Optimizer benchmarked against the solver you run today.