Robert Ambrogi Blog
At its Relativity Fest user conference in Chicago today, the e-discovery company Relativity announced the forthcoming release of Relativity aiR for Review, the first of a planned series of products that will use generative artificial intelligence to help legal professionals in their work.
Built on Microsoft Azure OpenAI Services, aiR for Review will enable user to more quickly locate data that is responsive to a discovery request, material relating to different legal issues, and key documents, Relativity said.
The announcement, which repeatedly emphasized the concept of “AI you can trust,” was made during the keynote address by Phil Saunders, chief executive officer, and Chris Brown, chief product officer.
To ensure users of the accuracy of the results it delivers, aiR for Review will allow users to view citations and rationale surrounding its recommendations and predictions.
aiR for Review will be released with limited availability by end of year, with expected general availability in the first half of 2024, Relativity said.
In an interview in advance of the keynote, Brown, the chief product officer, told me that, at this point in its development, aiR is supplemental to other forms of AI-assisted review, such as active learning or technology assisted review.
“Today, we view it as supplemental, and then we’ll see where the industry goes, just as TAR is still supplemental today for folks that still do manual review,” he said.
However, the advantage of aiR over those other forms of review is that it does not require training. Rather, review teams can establish context for the AI through the review protocol that the team is already creating.
“Part of the process is taking that protocol and making sure that it’s articulated in a way that can produce great results downstream from finding those responsive issues.”
Of the customers who have been testing it so far, he said, they have found good results across a variety of matters, ranging from jumpstarting an investigation in a small matter to quickly getting insights back to clients in large matters.
CEO Saunders opens the Relativity Fest keynote.
Two Other AI Products Released Today
Although aiR for Review will not be available until next year, Relativity announced two other AI-based products today that are both immediately available in RelativityOne, the company’s cloud-based e-discovery platform.
Both are updates of products originally developed by Text IQ, which Relativity acquired in 2021.
The first, Relativity Personal Information Detect, or PI Detect, is designed to improve the accuracy of personal information detection. The tool will help users more easily find and redact PI by using AI to identify patters and context that can help find PI and protected health information that other approaches might miss.
The other product, Relativity Data Breach Response, is designed to help organizations respond to data breaches more easily by assessing its impact and building a notification list.
PI Detect was formerly called Text IQ for Personal Information and Data Breach Response was formerly called Text IQ for Data Breach. From now on, the Text IQ brand will no longer be used in Relativity product names, the company said.
That includes Text IQ for Privilege, which is now rebranded as Relativity Privilege Identify, and which will be released for limited availability by end of year. It will use AI to quickly find privileged information with AI and reduce the risk of inadvertent disclosures.
In our pre-keynote interview, Brown said these product releases are much more than rebrandings, but re-engineerings to give them scalability, security and self-service ease of use.
“When we bought Text IQ, they were more of a tech-enabled services company,” Brown said. “We’ve been taking their technology, building it into a scalable RelativityOne product that’s integrated inside of Relativity One, and particularly making it partner enabled so that all of our partner channel and anybody, our enterprise customers as well, can take it to their customers and build practices around it.”
Managing AI Costs
A concern for anyone using generative AI products is the potential cost. Brown said RelativityOne’s management console addresses this issue by giving users transparency into their usage and costs. “We’re providing you with tools so you can see all the data in all of your workspaces and tenants and all the file types all around the world.”
The management console specifically allows users to track usage of new products that might have different cost models, he said.
“We’re giving you the user, if it’s not bundled in your base offering, we’re giving you usage data so you can predict before you use it and then after you use it know how much it will cost. And so we’re bringing that control to our partners so that they can then build their service models around that.”
Investing in AI
Beyond today’s announcements, Relativity said, it is continuing to invest in developing more AI-based solutions that will have a significant impact on customers, partners and the way they do work for their clients.
For one, it said that it is piloting a product that uses generative AI to develop a set of out-of-the-box models that help customers detect a range of compliance behaviors and identify documents where those behaviors are present.
The product will also allow users to leverage an intuitive interface to design their own reusable models and tailor them to their unique needs without having machine learning expertise. Relativity is also actively piloting generative AI to identify privilege.
The company said it will also pilot a generative AI solution to support case strategy and reduce the time and effort needed to analyze and organize case information.
Lastly, it is also developing and currently testing aiR Privilege Review, which Brown said would not be a replacement for Relativity Privilege Identify, but a boost. “We’re looking at a number of different experiments where perhaps the output of the deep-learning models in our privilege review can be inputs into generative AI and that can produce a better result, or vice versa.”
“Relativity has integrated AI into its products for over 15 years, and I’m proud that in a year where many felt immense pressure to rush and deploy the newest AI development for the sake of being the first to do it, Relativity stayed true to its steadfast approach to thoughtfully innovate with purpose,” CEO Saunders said of today’s announcements.
“The product updates shared in today’s keynote demonstrate Relativity’s commitment to mindfully and responsibly leverage the power of the latest generation of AI capabilities to build purpose-driven solutions, fit for the unique needs of our community.”