Ever since Relativity announced their new GenAI aiR suite of tools based on OpenAI's GPT-4 model,...
Relativity Translate: A case study
Summary
In 2022, Relativity launched Relativity Translate, an integrated AI (machine) translation function within RelativityOne. Although the software isn't a replacement for human translation — and AI translation isn't court-admissible — it still has its benefits.
This case study centres around a document review project that contained a large batch of documents in an unknown number of languages.
The challenge
A boutique law firm approached Altlaw for a document review that contained a document set with an unknown number of foreign languages.
With a tight disclosure deadline to meet, speed and accuracy were of utmost importance.
The solution
Using Relativity Translate, we're able to significantly reduce the time spent identifying and translating languages—all without sacrificing accuracy.
In this instance, we ran language analysis to determine how many languages were present and what percentage of each language existed in the document set.
Using the software we were able to simultaneously translate all documents into English after the language analysis was performed.
To ensure accuracy, we brought in reviewers who were familiar with the languages present during the review stage. Overall, our solution led to some excellent results.
Key gains
2,066 document units
53-minute translation time
1.54 seconds per 25,000 characters translated
The result
Before the introduction of Relativity Translate, this level of project would’ve taken days, meaning we saved the law firm a significant amount of time and allowed them to meet their deadline comfortably.
There were also substantial cost savings, as only documents marked as relevant were sent to an expensive human translator rather than the entire document set.
eDiscovery articles & news
Stay up to date with all things eDiscovery and information governance. Weekly updates ensure you’re always in the loop.