AI Agents for Industrial Scheduling

At Juna, I'm building scheduling solutions at the intersection of classical optimization and AI Agents. What excites me here is pushing the state of the art in two directions: automated heuristic generation (similar to what I explored with Algobench), and using agents to convert vague business requirements into well-defined optimization problems.

The core challenge in industrial scheduling is that every customer's problem is slightly different: different constraints, different objectives, different definitions of "good." Traditional optimization requires extensive customization for each use case. We're exploring how AI can accelerate this process, both by generating solution algorithms automatically and by helping translate business language into mathematical models.