IFORS 2026 Vienna, Austria

Events

Marion Technical Writer June 1, 2026

Hexaly is delighted to be the Headline Sponsor of IFORS 2026, the 24th Conference of the International Federation of Operational Research Societies. The event will take place from July 12 to 17, 2026, at the University of Vienna, Main Building, Vienna, Austria. You can access the conference program here.

Come visit the Hexaly booth and meet the Hexaly team throughout the IFORS 2026 conference. We will be happy to connect if you have questions, want to learn more about our optimization technology, or are curious about the latest features introduced in Hexaly 14.5. This new release delivers major performance improvements in routing, scheduling, and packing, as well as significant advances in nonlinear optimization. IFORS is also a great opportunity to discuss real-world optimization challenges, discover how leading companies use Hexaly, and explore career opportunities within our team in an informal setting. Below are the abstracts of Hexaly’s talk at IFORS 2026.

Hexaly Assistant: LLM-driven modeling for Hexaly
Fred Gardi, Co-Founder & CEO of Hexaly

Large language models (LLMs) enable the generation of optimization models from natural language. Prior work focuses on linear integer programs (NL4OPT benchmark), but Hexaly uses a non-linear set-based formalism well-suited to compactly model combinatorial problems such as vehicle routing, scheduling, and packing. We assess the capabilities of foundational LLMs in this context and show how results can be improved via prompt engineering and automatic syntax validation and correction in a closed-loop process integrated in Hexaly Studio.

Solving nonlinear continuous optimization problems with Hexaly
Julien Darlay, Co-Founder & Head Of Science of Hexaly

Hexaly is a model-and-run global optimization solver. In this talk, we present the heuristic and exact components used to address continuous nonlinear problems. Constraints and objectives are represented using an expression graph. Hexaly 14.5 introduces new components to improve the resolution of such problems, including structure detection, a new modular interior-point algorithm, and a branch-and-reduce algorithm with cut generation. This new version has been benchmarked against open-source implementations on three benchmarks: pooling, portfolio optimization, and the CUTEst library.

Branch-cut-price methods for routing problems in Hexaly
Noémie Cartier Ph.D., Co- Optimization Scientist

Hexaly is a model & run mathematical optimization solver based on exact and heuristic techniques. Its modeling formalism includes list variables that can be used, for example, to model routes in routing problems in a compact way that is well-suited to a heuristic search. This talk will focus on how Hexaly automatically reformulate such structures and uses branch-cut-price techniques to generate lower bounds on a wide variety of routing problems. We will present the mathematical tools used and some benchmarks on some classical routing problem families.

We look forward to connecting with the Operations Research community at the 2026 IFORS conference in Vienna, Austria.

Feel free to explore our upcoming conferences, events, and webinars.