Solver Workbench

DocsBook a demo

Solver Workbench

From code to running solver, without the plumbing

Start from a template or connect your repository. The platform builds, validates and runs your solver across quantum, GPU and classical providers.

Book a demoRead the docs
Solutions explorer: solvers matched to a use case, mapped to quantum, quantum-inspired and classical providers, with potential, complexity and cost

Solver Workbench is where developers and researchers build quantum, quantum-inspired and classical solvers on the QCentroid QuantumOps platform.

Start from a provider template or connect a Git repository; QCentroid checks the code contract, builds the solver and runs it on any supported backend. Your algorithms become validated, versioned solvers the whole organisation can benchmark and reuse.

  • Telefónica
  • Mahou San Miguel
  • MasOrange
  • GSMA
  • Moeve
  • Moody's
  • Sabadell BStartup
  • BiQain
  • BackFund
  • QAI Ventures
  • CDL
  • Telefónica
  • Mahou San Miguel
  • MasOrange
  • GSMA
  • Moeve
  • Moody's
  • Sabadell BStartup
  • BiQain
  • BackFund
  • QAI Ventures
  • CDL

What is Solver Workbench?

Start fast

Ready-made templates for each provider and SDK.

Keep your workflow

Connect Git; the platform builds and runs your code.

Fail early, not late

Code checked before the build, not after a failed job.

Features

What can you do with Solver Workbench?

Six tools to take an algorithm from code to a solver others can run: pick what already works, start from a template, generate a first draft, connect your repository, validate the build and publish it.

Solutions Explorer

Browse the solvers available for your problem and see how they perform across quantum, quantum-inspired and classical backends before you choose.

  • Solvers matched to your problem type
  • Performance compared across 10+ backends
  • Pick the right solver and hardware for each job

Why it mattersTeams start from what already works instead of rebuilding from scratch, and choose on evidence, not vendor claims.

Solutions explorer: solvers matched to a use case, mapped to quantum, quantum-inspired and classical providers, with potential, complexity and cost
Solutions explorer: solvers matched to a use case, mapped to quantum, quantum-inspired and classical providers, with potential, complexity and cost

Solver Templates & Runtime

Start from a ready-made template for your provider and SDK. A versioned runner handles libraries and hardware calls.

  • Templates per provider and SDK
  • One entry point: a single run() function
  • Versioned runtime for reproducible runs

Why it mattersDevelopers write the algorithm, not the integration code.

Add a new solver: choose the implementation method (code repository, container repository, external worker or external service)
Add a new solver: choose the implementation method (code repository, container repository, external worker or external service)

Code Generator

Generate a first version of your solver code from the use case formulation, ready to review, edit and run on the platform.

  • Solver code drafted from the formulation
  • Follows the platform's solver template and entry point
  • Developers review and own every line

Why it mattersLess boilerplate, faster first runs, and expert developers stay in control of the final code.

Coming soon: this feature is in development.

Solver code approach: generate the code, pseudocode and flowchart of a solver, with push to repository and download
Solver code approach: generate the code, pseudocode and flowchart of a solver, with push to repository and download

Repository-Connected Solvers

Connect a Git repository and the platform builds and runs your solver. Push a change, rebuild, run.

  • Connect your Git repository in a few clicks
  • Automatic build from your code and dependencies
  • Rebuild after every change

Why it mattersKeep your usual tools and version control. No copying code into a web form.

Connected Git repositories with their status, number of solvers and the time of the last build
Connected Git repositories with their status, number of solvers and the time of the last build

Build-Time Code Validation

The solver's code contract is checked before the build, so errors show up in seconds, not after a failed job.

  • Entry point and signature checked before build
  • Clear errors that point to the fix
  • No compute wasted on broken builds

Why it mattersFailed jobs on real hardware cost time and money. Catch errors before they run.

Repository page of a connected solver with its builds history and the entrypoint check (qcentroid.py, run() and signature) passed
Repository page of a connected solver with its builds history and the entrypoint check (qcentroid.py, run() and signature) passed

Solver Catalog

Publish solvers as documented, branded entries others can find and run, with control over who sees them.

  • Documentation shipped with every solver
  • Branding for partner and vendor solvers
  • Visibility: private, organisation or public

Why it mattersOne team's work becomes a reusable asset across your organisation or ecosystem.

Organisation solver catalogue with type, lifecycle stage, use case and estimated cost of each solver
Organisation solver catalogue with type, lifecycle stage, use case and estimated cost of each solver

Questions about Solver Workbench

How do I deploy a quantum algorithm from my own Git repository?

In QCentroid Solver Workbench, you connect a Git repository and the platform builds the solver from your code and dependencies. After a change you rebuild and run, keeping your usual tools and version control.

Do I need to write integration code for each quantum provider?

No. QCentroid provides solver templates per provider and SDK, and a versioned runner handles libraries and hardware calls. Developers implement a single run() function and focus on the algorithm.

How does QCentroid avoid failed jobs on expensive hardware?

Solver Workbench checks each solver's entry point and signature before the build, with clear errors that point to the fix. Broken code is caught in seconds instead of after a failed run on a QPU or GPU.

Solver Workbench

Put your algorithms to work.

Template, repository, validated build, catalog.

Book a demoRead the docs
Organisation solver catalogue with type, lifecycle stage, use case and estimated cost of each solver