United StatesFinance and insurance2017 – ongoingJPMorgan Chase
COiN contract intelligence for commercial loan agreements
JPMorgan's COiN software reviews loan agreements in seconds, replacing 360,000 hours of work a year
Problem
JPMorgan's lawyers and loan officers spent about 360,000 hours a year interpreting around 12,000 new commercial credit agreements. Misreadings of these contracts led to loan-servicing errors.
Approach
The bank built COiN (Contract Intelligence), a machine learning system that reads the agreements and extracts the relevant terms, learning patterns from the data it ingests. By early 2017 it was in use on commercial loan agreements, as part of a wider push to automate work with technology.
Results
Takeaway
Standardised documents reviewed in high volume gave both time savings and fewer human reading errors. The figures came from the bank in February 2017.
Categories
- Size
- Large enterprise
- Industry
- Finance and insurance
- Business function
- Finance, risk and legal
- Innovation type
- Process innovation
- AI technology
- Predictive analytics and ML
- Adoption
- Built in-house
- Evidence
- Company's own claim
Sources
Source summaries
The ABA Journal, citing Bloomberg on 28 February 2017, reports that JPMorgan's software reviews documents in seconds and handles work that took lawyers and loan officers 360,000 hours a year. The bank applies it to about 12,000 new contracts a year and says it has helped reduce loan-servicing mistakes caused by human error.
Futurism reports that COiN cut about 360,000 hours of yearly work on commercial loan agreements to seconds. The system's designers linked loan-servicing errors to human misreading of some 12,000 new wholesale contracts a year.
Full details are in the original sources.
