What if anyone could talk to a solver?
That's the question I keep coming back to. It´s a pain point of mine. Then, I have been studying the intersection of LLMs and mathematical optimization seeking to figuring out how natural language can become the front door to powerful decision-making engines.
AI that decides, and accessible to non-operations research professionals.
- Decision Support Prototypes: Projects where you describe your scenario in plain English and get back an optimal answer.
- Skills & Commands: Custom workflows I've built to stay productive with GenAI.
- My Learnings: All my experiments, structures I have tested, and what has worked for me when combining LLMs and solvers.
My most recent prototype is Full-Truckload (FTL) Planning with LangGraph, in which I combine previous learning with analysis of the LangGraph framework.