White paper features results and testimonials from leading law firms
NEW YORK–(BUSINESS WIRE)–Syllo, the litigation workspace for the AI era, and a group of seasoned litigators and eDiscovery professionals today announced the release of a white paper entitled Agentic AI Document Review Is Transformative for Complex Litigation. The paper is a collaboration between Syllo, Professor Jamie Callan of the Language Technologies Institute at Carnegie Mellon University, and 25 litigators and eDiscovery practitioners from seven elite law firms. The publication for the first time describes Sylloās novel approach of using agentic AI to automate document review in complex investigations and litigations. It also provides validation studies describing how elite litigators have used Sylloās agentic review system in live litigations and metrics confirming that the system far exceeds performance benchmarks of prior forms of technology-assisted review.
āFor GenAI to be a true difference-maker in complex litigation, litigation teams need to be able to deploy it accurately and cost-effectively at scale to review tens of thousands, hundreds of thousands, and millions of documents for highly nuanced issues,ā said Jeffrey Chivers, Sylloās co-founder and CEO. āWe are excited to share more about the solution we built for this and the impact it is having in complex litigations.ā
Sylloās agentic document review system represents a significant advancement in the state of the art. The system orchestrates multiple LLMs performing distinct roles to autonomously complete detailed analyses across large and complex litigation datasets under the guidance of subject matter experts. āThe document tagging solution architected by the Syllo team is a novel way to solve the problem of reviewing large data sets and the results it generates represent a notable advancement in the field,ā said Professor Callan.
Twenty-five attorneys and eDiscovery practitioners contributed to the white paper from the law firms of Ballard Spahr LLP, Mayer Brown LLP, Nixon Peabody LLP, Outten & Golden LLP, Pillsbury Winthrop Shaw Pittman LLP, Quinn Emanuel Urquhart & Sullivan, LLP, and Royer Cooper Cohen Braunfeld LLC.
The paper details the breakthrough performance that these law firms obtained using Sylloās agentic AI system. Syllo has achieved an average estimated recall of 97.8 percent and an average estimated precision of 79.7 percent in the last ten completed responsiveness reviews.
Additional highlights include:
- An elite litigation firmās use of Syllo in a complex commercial litigation to apply over 25 issue codes to 100,000 documents while obtaining estimated Precision of 95.56% and estimated Recall of 99.4%.
- More than four automated document reviews in which Syllo achieved an estimated Recall of 100%.
- Numerous validation studies detailing how Syllo reviews large document sets faster and more cost-efficiently than managed review teams and older technology-assisted review technologies, while unlocking for litigation teams unprecedented speed and control over the litigation record.
āWe are extremely grateful to the litigators and eDiscovery professionals whose expertise helped align the development of Sylloās agentic AI review system and who contributed to the paper so that the advantages they have gained from leveraging the system can be more widely understood by the litigation and eDiscovery industry,ā Chivers said.
Technical contributors to the paper included Sylloās Head of Machine Learning, Pei-Lun Tai, Sylloās CEO Jeffrey Chivers, members of Sylloās engineering team, and Professor Jamie Callan of the Language Technologies Institute at Carnegie Mellon.
Read the full whitepaper here.
About Syllo
Syllo is an AI-powered litigation platform that enables legal teams to safely and securely harness the power of language models throughout the litigation life cycle. Founded in 2019 by a team of litigators and engineers, Sylloās litigation platform provides a competitive edge to case teams, practice groups and law firms.
Contacts
Eric Wall
Executive Vice President
Syllo
415-613-7927
[email protected]
Bill Hughes
772-268-6828
[email protected]