Fileread User Webinar 2025

How Two Law Firms Are Using AI to Transform Legal Discovery: Insights from Fileread Power Users

Generative AI is reshaping legal workflows, but how are firms actually implementing these tools in practice?

In the recent Fileread user webinar, hosted by our VP of Growth Justin Brownstone, two experienced legal professionals shared their real-world experiences using AI-powered document review in high-stakes litigation.

Meet the Users

Anna Kovaleska, an associate at Fields Han & Cunniff, represents state attorneys general and tribal nations in high-impact civil litigation. As a small firm of six lawyers going up against corporate defendants, her team constantly asks: "How do we punch above our weight?"

Kate Jansons Johns manages knowledge and practice support services at Nutter, a 150-attorney Boston firm with about 40 in litigation. They keep most e-discovery work in-house using Relativity One and focus heavily on maximizing their technology ROI.

Watch

The Problem They Sought to Solve

Kate's journey with Fileread began in 2021 when she wanted to create "a Google-like search experience" for attorneys using Relativity. Traditional Boolean searches and conceptual analytics weren't delivering the intuitive experience lawyers needed.

"I just wanted to have a Google type experience," Kate explained. "We were using Fileread before ChatGPT became a thing."

For Anna's smaller firm, the challenge was efficiency: delivering the same caliber of work with greater efficiency and less manpower while building strong factual records to support their legal theories.

Real-World Applications

Quick Intelligence Gathering

Anna shared a practical example: when tasked with identifying pharmacies for a production request, she used Fileread to search pharmacy names one at a time. The tool didn't just return associated documents—it surfaced additional intelligence like DEA numbers, alternate business names, misspellings, and owner information.

"In a short time period, I was able to collect a much more robust set of data and get to the facts quicker," Anna noted.

The "Sandwich Approach" to Document Review

Kate has developed an innovative three-stage workflow combining Fileread with Relativity's AiR for Review:

1. Start with Fileread: Attorneys interrogate documents to understand what's there without spending excessive time

2. Move to Relativity AiR for Review: Use insights to develop better prompts for predictive coding and issue tagging

3. Return to Fileread: Generate fact memos summarizing responsive and non-responsive document buckets
   
   

"The idea is that an associate starting their review will have some kind of summary about what's in the documents and can prioritize what they're reviewing in a more mindful way," Kate explained.

Overcoming Adoption Challenges

Making It Discoverable

Kate's strategy for driving adoption: make the tool impossible to ignore. She places Fileread as the second tab in every Relativity workspace, sometimes calling it "AI Assisted Searching and ChatGPT for Relativity."

"I'm being kind of passive aggressive and putting it somewhere where it's easy for attorneys to stumble upon," she admits. "You can bring them to the water, but you can't make them drink."

Building Trust Through Transparency

Both users emphasized the importance of Fileread's citation engine for building attorney confidence. The tool cites directly to source documents within Relativity, making verification straightforward.

"The fact that the analysis will cite right to the document in Relativity" has been crucial for adoption, Kate noted. Anna added that she thinks of generative AI "like an intern or assistant that's very new" — helpful for initial analysis, but attorneys must always verify the work.

Tips for New Users

Anna's advice: Treat Fileread like ChatGPT. If you don't get what you want immediately, try rephrasing your question. "Sometimes the human doesn't understand you," she noted, recommending experimentation with variations like "identify emails instead of documents" or "identify spreadsheets."

Kate's guidance: Ask for specific output formats. Don't just accept a string of information—request charts, groupings, or specific organizational structures.

Anna offered a memorable analogy: "Think of Fileread as a person who has all the information in the document set. You can ask it anything as many times as you want, and it won't get tired."

Looking Forward

Both users see AI as essential for managing ever-increasing data volumes. As Kate put it: "There's just so much data now, and our brains aren't getting any bigger. We have to turn to these AI tools to manage this data and keep this business afloat."

The key insight from both firms: successful AI adoption requires making tools accessible, building verification processes, and allowing attorneys to discover the value through hands-on experimentation. As the legal industry grapples with data explosion, these practical approaches offer a roadmap for meaningful AI integration.

To learn more about Fileread, contact Justin at justin@fileread.com