BNY Mellon

BNY Mellon’s legal team manually reviewed vendor agreements in a multi-step workflow that required cross-referencing prior contracts, validating global regulations, benchmarking pricing, and flagging risks. Each review took about 4 hours, and the volume—3,000+ vendor agreements annually—previously required multiple full-time paralegals.

BNY Mellon deployed a Contract Review Assistant on its Eliza platform as part of a broader agentic AI program (20,000 Empowered Builders and 130+ Digital Employees). The workflow uploads a contract, extracts terms, cross-references 3,000+ prior agreements, validates against global regulations, benchmarks pricing, flags risks, and notifies the legal team via Microsoft Teams while keeping attorney judgment in the loop.

Legal contract review time fell 75%, from 4 hours to 1 hour, while quadrupling throughput. The assistant now processes 3,000+ vendor agreements annually. The Eliza platform runs 125+ live use cases in production across the 52,000-person workforce.

• • Agentic contract review that cross-references prior agreements and regulations can cut review time by 75% while preserving attorney judgment
• • Measurable ROI (hours saved and throughput gains) is clearer when agents execute end-to-end workflows rather than answering chat prompts
• • Large productivity gains on data-heavy legal grunt work may reduce demand for junior/paralegal capacity even when senior judgment roles are augmented