How Troutman eMerge Stays Ahead: A Q&A on AI-Powered Legal Discovery
Troutman eMerge, the eDiscovery and legal technology arm of Troutman Pepper Locke, has been an early and deliberate adopter of AI in legal practice. We sat down with their team to hear how Fileread is changing the way they work.
Customer
Troutman eMerge
Innovators
Alison Grounds
Founding & Managing Partner
Jason Lichter
Principal
Mike Franke
Counsel
Antonio Avant
Director of Legal Technology
Website
troutmanemerge.com
"Be bold with your questions. Attorneys are used to tools that do rote things. Fileread goes deeper, and you'll often be surprised by what it finds when you actually ask what you want to know."
Key Highlights
Fileread enables document review that is more accurate and consistent than human review, supported by an ISO-certified AI governance framework.
Attorneys can ask natural language questions instead of relying on keywords, allowing them to instantly surface key themes, facts, and timelines.
Fileread works directly inside Relativity, eliminating workflow disruption and the need to switch tools.
The team is now able to identify pattern shifts across documents, uncovering signals that would be difficult to detect through manual review.
How does eMerge think about generative AI in legal work?
Alison Grounds, Founder and Managing Partner: If you're not using generative AI to assist in rendering legal services, especially when large document analysis is involved, you're shortchanging yourself and your client. Early on, we found that AI applied to document review was more accurate and consistent than human review. Once you have that evidence before you, it would be a real mistake not to take advantage of it. We just received an ISO certification for our AI governance framework, because deploying these tools responsibly is just as important as deploying them at all.
Where does Fileread fit within your broader technology approach?
Alison: Fileread is one of many technology tools we deploy to make sense of information quickly. We're not just licensing anything we can get our hands on. We're licensing tools from trusted partners that we've been able to validate and have repeatable results from. Our use of generative AI has become essential to how we propose working with clients. It's one of the reasons they choose us. We've been early adopters, and we're combining the right people, process and technology to deliver a better result.
What specific problem did Fileread solve for your team?
Antonio Avant, Director of Legal Technology: We needed a tool that would sit on top of our Relativity platform and let us ask real questions of a document set in plain language. Not just run search terms, but actually have a conversation with the data. We can identify key themes, pull out key facts, build a chronology and help the case team understand the chain of events in a matter before doing a full review.
How does it fit into the broader workflow?
Jason Lichter, Principal: Fileread doesn't replace Relativity, it extends it. There's no context switching, no new platform to learn. It's a click away inside the environment our attorneys already know. Post-production review is the most time sensitive phase of litigation. The other side may produce hundreds of thousands of documents with depositions days away. Fileread is what allows us to find the needles in that haystack quickly and accurately.
Walk us through how you actually use it in a case.
Mike Frankel, Counsel: I use it at almost every stage. In early case assessment, I start with exploratory questions: who are the key players, what's the terminology being used, what decisions were made and when. During review, I'll use Fileread to write better prompts for our other AI tools, because I have a clearer picture of how people in the documents are actually referring to things. For deposition prep, we feed in the complaint, the interrogatories and the key documents tied to a specific witness and ask Fileread to map the timeline and identify the most likely lines of inquiry.
Any results that stood out?
Jason: Just recently, a colleague used Fileread to accurately anticipate the topics opposing counsel raised in a deposition before walking in the door. That kind of preparation used to take days. Fileread made it possible in hours.
Mike: We had a trade secret case where we needed to detect whether former employees had been conspiring to leave and take proprietary data. Fileread found not just what people said, but what they stopped saying. Pattern changes subtle enough to miss document by document, but visible when you look across the full set.
What would you tell a team still on the fence?
Mike: Be bold with your questions. Attorneys are used to tools that do rote things. Fileread goes deeper, and you'll often be surprised by what it finds when you actually ask what you want to know.
Alison: The teams who figure out how to combine the right people, the right process, and the right technology are the ones who will win. Fileread is part of that combination.