AI that auto-builds large-scale optimization models (LEAN-LLM-OPT)
Big business decisions often rely on complex optimization models, but building them is slow and manual. Meet LEAN-LLM-OPT, a lightweight, multi-agent AI that auto-formulates large-scale optimization models from a plain-English problem description and datasets.
How it works: two planner agents design a step-by-step workflow for similar problems; a builder agent follows it to produce the final formulation, while routine data handling is offloaded to tools. The result: the AI focuses on tough modeling choices instead of bookkeeping.
- Why it matters: faster prototyping, fewer errors, and more consistent modeling.
- Performance: Strong results in extensive simulations; competitive with state-of-the-art. In a Singapore Airlines choice-based revenue management case, it delivered leading performance across scenarios.
- Models used: GPT-4.1 and the open-source gpt-oss-20B.
- New resources: Two benchmarks (Large-Scale-OR and Air-NRM) plus code and data: https://github.com/CoraLiang01/lean-llm-opt
Paper: https://arxiv.org/abs/2601.09635v1
Paper: https://arxiv.org/abs/2601.09635v1
Register: https://www.AiFeta.com
AI OperationsResearch Optimization LLM Automation RevenueManagement OpenSource Benchmark
