Machine learning software reduces lawyer effort by more than 50% and speeds up delivery to the client
A global strategic client instructed our Paris litigation team to work on a matter that involved reviewing a set of employment contracts to identify relevant information and provide advice in relation to our findings.
CASE STUDY | CONTINUOUS IMPROVEMENT, KIRA
The lawyers were faced with an extremely short deadline to review over 1,000 employment contracts. Most of the contracts were written in French, with a sizable minority in German. Traditionally, a lawyer would coordinate the review exercise with colleagues via email. Contracts would be sent to Clifford Chance lawyers in the relevant countries. The lawyers would then review all 1,000 contracts and send back their findings to be collated into a report for the client.
Due to the tight deadline, the litigation lawyers worked with the Best Delivery team to work out how to speed up delivery.
The team ran a short workshop to understand the process and the output required. It was decided to use Kira, a piece of machine learning software that can identify and extract key clauses from contracts and other documents in a fraction of the time a person would require to carry out the same task.
The Best Delivery team, working with the lawyers and their secretaries, created a list of key words and phrases that would indicate that a contract needed further scrutiny. All 1,000 contracts were uploaded into Kira and searched for the key terms.
Kira was able to identify 550 contracts that could be excluded from further review as they did not contain any of the identified words or phrases. This meant that our lawyers could concentrate their detailed manual review on the remaining 450 contracts. The online Kira platform allowed us to quickly and efficiently assign the contracts to the relevant French and German speaking lawyers. The manual review of the reduced sample was then completed online.
Our team estimate that the lawyers were able to deliver the final report to the client one week earlier than would have been possible using traditional methods.