Automatically Review Leases with NLP
A “Big 4” Accounting firm needed to speed up the review of their client’s lease agreements, extracting specific clauses and rates
They wanted to categorize leases based on which contracts held certain provisions or attributes. Manually reviewing thousands of thirty-page contracts is time consuming, costly, and prone to errors. Skilled reviewers’ fatigue after a few hours of reading similarly worded documents. An option they investigated was outsourcing the review to less expensive workers overseas. However, time and accuracy were still an issue.
The Lymba Solution
Accurately and automatically review leases while extracting key information. The automatically extracted lease information is stored in a RDBMS database for querying and report generation.
Process
Properties and terms can be set for multiple purposes; custom named entity types for specific use; domain-specific semantic relations for each case. A user simply uploads files, our pipeline captures the exact meaning of contract provisions and extract contract paragraphs matching that meaning.
Outcome
Based on a sample size of 100 leases, we saved 400 man-hours of review. Using a modest billing rate of $200/h, this firm saved $80,000 to review those 100 same leases. This is an ROI of over 33%. And it is scalable!