Constraint Programming (CP) is a powerful paradigm for solving complex combinatorial problems where traditional algorithms may be inefficient or infeasible, and generative AI doesn't quite handle the complexity.
CP allows us to declare constraints that define the rules of the problem, letting a solver explore the solution space intelligently.

This approach is particularly effective in domains where LLMs currently struggle, such as scheduling, routing, resource allocation, and other optimization tasks that can be used in many real-world applications.

In this talk, Jakob will introduce the core ideas of Constraint Programming and how this can be combined with generative AI, showing several demos using Google’s open-source OR-Tools framework