Streamlining Industrial Contract Management with Retrieval-Augmented LLMs

Contract management involves reviewing and negotiating provisions, individual clauses that define rights, obligations, and terms of agreement. During this process, revisions to provisions are proposed and iteratively refined, some of which may introduce legal, operational, or business risks and therefore need to be identified before they delay negotiation or lead to unfavorable terms. Automating this workflow is challenging because organizations often have large collections of legacy contracts, but only limited labeled examples explaining which revisions were ultimately acceptable and why. In this paper, we present a modular framework designed to streamline contract management through a retrieval-augmented generation (RAG) pipeline. Our system integrates synthetic data generation, semantic clause retrieval, acceptability classification, and reward-based alignment to flag problematic revisions and generate improved alternatives. Developed and evaluated in collaboration with an industry partner, our system achieves over 80\% accuracy in identifying problematic revisions and over 80\% successful optimization of unacceptable revisions, demonstrating strong performance under real-world, low-resource conditions. These results suggest that legacy contract repositories can be converted into practical decision-support tools for accelerating contract revision workflows while preserving expert oversight.