Skip to content
Skip to Content
Research & open source

We don't just use solvers. We build the bridges.

NexOR's optimisation team contributes to JuMP, the Julia ecosystem's modelling language for mathematical optimisation. Several of the connectors between JuMP and the solvers underneath our VRP engine live in our public repos. Run NexOR knowing the math underneath is auditable.

Visit the GitHub organisation What is JuMP?

github.com/
NexOR-Optimization
  • 4 Public Julia packages
  • EU Belgium-based
  • MIT Open licences
JuMP, in one paragraph

An open-source modelling language for mathematical optimisation.

JuMP lets researchers describe an optimisation problem, routing, scheduling, network flow, mixed-integer programming, in close-to-mathematics syntax, then solve it with any of thirty-plus solver backends behind a single interface.

It is the de-facto standard in the Julia ecosystem and the backbone of optimisation research across European universities, including UCLouvain, where parts of our optimisation team trained.

  • Created at MIT, used in academic research worldwide
  • One model, thirty-plus solver backends
  • MIT-licensed, community-maintained

jump.dev

Built on
What we ship to the community

Solver bridges for the Julia ecosystem.

These are some of the connectors we maintain. The bridges between JuMP's modelling language and the solvers behind the NexOR VRP engine. Open source by design.

Julia 5
Hexaly.jl

JuMP interface to Hexaly, a high-performance commercial constraint-programming and metaheuristics solver. Lets Julia models drive Hexaly directly.

View on GitHub
Julia 3
MaxiCP.jl

JuMP interface to MaxiCP, an academic constraint programming solver maintained at UCLouvain. Open-source all the way down.

View on GitHub
Julia 0
Vroom.jl

JuMP interface to VROOM, a popular open-source Vehicle Routing solver. Brings VROOM into the Julia/JuMP optimisation toolbox.

View on GitHub
Julia 2
ContractionHierarchies.jl

Fast shortest-path computation on OpenStreetMap graphs using contraction hierarchies. The geo-routing primitive underneath our distance and travel-time matrices.

View on GitHub
Why this matters when you buy a TMS

Open math beats black-box math.

Four reasons NexOR's open-source posture is a feature, not a footnote.

01
Auditable models

The math behind your routes is in code anyone can read. No proprietary black box deciding what your trucks do.

02
Vetted by research

JuMP and its solvers are used in operations-research labs worldwide. The models we ship are scrutinised by the global community.

03
No lock-in

If we vanish tomorrow, the solver bridges stay alive on GitHub. You keep the math, you keep the code, you keep running.

04
Customer-shaped

A solver that doesn't model your reality is just slow software. We extend the open packages when a customer constraint isn't already supported.

The team behind the math

Operations research, in the cab.

NexOR's optimisation work is grafted into the Julia language ecosystem. Members of the team are active contributors to JuMP itself, shipping pull requests against the core packages, maintaining solver interfaces, and presenting at the JuMP-dev workshop. The bridges above are not a side project. They're how we work.

  • Active contributors to the JuMP core packages
  • Talks at the JuMP-dev workshop
  • Solver interfaces published as open source on the org
  • Daily product work and research run on the same code

Meet the team

github.com/NexOR-Optimization
Solver bridges, geo-routing primitives, and the NexOR platform itself.
Browse the org

Got an optimisation problem we should look at?

Send us the constraints you can't model in your current tool. We come back with what's already in JuMP and what we'd build.

Talk to the optimisation team