AI-Enabled Contracts Comparison

Compares two credit agreements section by section, using structure-aware vector indexing and two-stage retrieval, then rates each discrepancy's severity and risk direction. By Partha Sarkar Project page: https://towardsdatascience.com/shipai/projects/contract-comparison ShipAI is a Towards Data Science project. You can learn more about it here: https://towardsdatascience.com/shipai/about Key takeaways. 1. Given a criterion, it takes 10 relevant sections from the first contract and matches up to three from the second. 2. PDFs become Markdown and a skeleton tree, so chunks never cross sections; vector recall plus LLM ranking does the matching. 3. Each pair gets a severity, risk direction and analysis; Texas Roadhouse's missing setoff sharing was flagged significant. Chapters: 0:00 Why contract comparison is hard 0:40 Emerson vs Texas Roadhouse 1:42 Choosing the criteria 2:22 Indexing and matching sections 3:22 Reading one comparison 5:26 A financial covenants report 7:17 The architecture