
Multi-capability legal tech group, Epiq, is upping the ante on its AI offering, especially for eDiscovery, with the launch of ‘Epiq AI Labs’. It is also rolling out its AI Discovery Assistant, which they say is ‘90% faster than TAR or linear review’.
The AI Discovery Assistant, which is available through its Service Cloud ‘automates more than 80% of traditional eDiscovery processes and completes reviews up to 90% faster than TAR or linear review, with up to 4,000x throughput, including large and complex datasets’, they explained.
The moves follow the acquisition of Laer AI in 2024, which along with its AI tech saw founders Igor Labutov and Bishan Yang, who have doctorate degrees in Computer Engineering and Computer Science from Cornell University, join Epiq.
See In-Depth AL Interview below that covers this move and how Epiq is achieving these increased performance metrics.
As to what Epiq AI Labs will do, the company said that it will be a ‘community of technologists, academics, and industry leaders focused on building foundational technology’.
It will seek to develop new AI technologies for the most ‘pressing problems in large-scale data understanding and legal service delivery’.
While in terms of what the AI Assistant can do, it simultaneously reviews data for issues, privilege, and responsiveness, using prompts and models automatically created from review protocols.
‘Without repeating workflows, the software connects information across data sets to form insights that could not have been discovered using natural language search within individual documents. Clients gain speed to knowledge, plus the ability to analyze up to 500,000 documents per hour,’ they explained.
They also added that the tech and services offering will be available on a ‘predictable pricing’ basis, which will help parties to better understand their costs.
And in relation to the human side of things, Epiq is offering its own Review Services so inhouse teams, or law firms, can ‘collaborate with Epiq AI consultants, review managers, and teams to expedite review using’ its AI tech.
The company was understandably excited to share the views of a partner from elite New York firm Sullivan & Cromwell, who had used the AI assistant.
Matthew Schwartz, a partner at Sullivan & Cromwell, said: ‘It significantly reduces the volume of documents we need to have first-level reviewed and the time it takes to do so. The high confidence scores that we have seen when using the program, as well as its ability to extract information quickly from the documents, allows our teams to work faster and more efficiently.’
All well and good. It’s interesting to see that although we often focus on how genAI is impacting the world of contracting, it’s also having a notable impact on eDiscovery as well.
Another aspect worth noting here is the combined use of managed services. Epiq is not pretending AI will do the whole thing, even if eDiscovery reviews can now be done a lot faster, and therefore perhaps need less people per matter.
More broadly, this and other improvements in eDiscovery review, raise the question of whether the total number of staff needed for such work will decrease, whether they are working at ALSPs that focus on such work, or within law firms. This site has not seen specific data on such demographic changes, but it seems logical that one needs fewer manual reviewers in total across the sector if so much more of that work can be automated now.
Eric Crawley, Senior Vice President of Legal Solutions at Epiq, concluded: ‘[These initiatives] build on years of data, analytics, and AI development.
‘With Microsoft and Amazon (AWS) capabilities, access to proprietary Epiq AI technologies, services, and applications, including custom AI application development, our clients can leverage AI across their entire legal services management focus.’
See more on his thoughts below.
—

AL Interview with Eric Crawley, Senior Vice President of Legal Solutions at Epiq:
- Epiq has worked in this field for some time, what has changed that allows these advances? Is it primarily the arrival of genAI?
In hiring experts like Igor Labutov and Bishan Yang, and acquiring the rights to the technology they’ve built, Epiq recognizes an opportunity to recast the discovery process and deliver additional value to our clients that other processes and workflows cannot. We feel it is important to deliver that value to the market now, creating new solutions for the future of legal work. We believe this will differentiate Epiq from its competitors and take a leading role in educating the market and driving the adoption of AI solutions.
- Can you please explain how you achieve such improvements e.g. 80% automation. Maybe several bullet points that set out the process?
Epiq AI Discovery Assistant automatically creates a structured knowledge layer, prompts and models, using an agentic approach to select the right technology or the right model for the right task. Attorneys can run an AI-based review using a workflow they’re familiar with, creating and submitting the same review protocol they’d give to human reviewers to our system to perform a complete review.
- What changes does this have for the price of such Ediscovery projects for law firms, and the manpower needed to run such projects?
Epiq AI Discovery Assistant is available on a per-matter basis with predictable pricing that includes unlimited prompts and protocols without triggering additional fees.
- Could such tech be used to help with transactions and also large-scale contracting needs?
Yes, we’ve identified a number of use cases and applications for this technology. They are built into our roadmap for the coming year and beyond. Stay tuned for more details…
- Plus, why have the experts from Cornell joined?
Epiq, as well as Labutov and Yang, has a long-standing relationship with Cornell. Cornell wanted to form Epiq AI Labs with us on the shared belief that there’s a need for the industry to better understand the practical use of AI in law. Cornell and other organizations, to be named later, look forward to helping us educate the market and collaborate on new technology.
—
(Main pic: the Epiq Service cloud.)