JPMorgan Chase processed thousands of commercial loan agreements annually, each requiring detailed legal review to extract key terms, identify obligations, and assess risks. This manual document review consumed approximately 360,000 hours of lawyer and loan officer time per year, creating an expensive, error-prone bottleneck that delayed deal closings. The heterogeneous document formats — spanning scanned PDFs, Word files, and legacy image-based archives — demanded robust optical character recognition and semantic parsing capabilities.
With a lending portfolio spanning dozens of jurisdictions, each contract incorporated distinct legal conventions, terminology, and structural formats that compounded the semantic-understanding challenge. The bank needed an automated extraction system whose outputs were attestable in regulatory examinations, imposing stringent audit-trail requirements on every algorithmic annotation while reconciling extracted data with downstream risk models.
JPMorgan developed COiN (Contract Intelligence), a machine-learning system that uses natural language processing to analyse commercial loan agreements and extract relevant data points automatically. Launched in June 2017, COiN was trained on thousands of historical contracts, learning to identify key clauses, flag non-standard terms, and extract structured data for downstream credit-risk systems.
The platform runs on JPMorgan's private cloud infrastructure and integrates directly with the bank's loan-processing workflow, automatically analysing contracts and surfacing findings for human review. A confidence-tiered workflow routes high-certainty extractions for auto-acceptance while flagging lower-confidence outputs for specialist legal reviewers, with an active-learning feedback loop continuously incorporating corrections into model retraining.
COiN formed part of JPMorgan's broader AI transformation, which by 2025 included the LLM Suite generative AI platform adopted by over 200,000 employees, over 450 AI use cases in production, and an annual technology budget of $17 billion — the largest of any financial institution.
This is an industry case study based on publicly available information. JPMorgan Chase is not a Pertama Partners client.
We help enterprises across Southeast Asia design and deliver AI transformation programs. Let’s talk about what’s possible for your team.
