MIDAS: Multi-LLM Iterative Data-Adaptive Summarization

MIDAS implements a multi-LLM framework that utilizes iterative critique-driven refinement and data-driven pattern learning to automate summarization. The system moves beyond static prompts by personalizing outputs to match domain-specific guidelines and organizational formats through continuous, use-case-specific adaptation.