ReferU.AI is an AI-powered platform that connects individuals and businesses with attorneys who have verified experience grounded in court records instead of marketing proxies. ReferU.AI, founded by Joel Geddis, applies agentic AI and large-scale court data to solve major problems in the legal system: unequal access to information and quality representation.

The idea for ReferU.AI came from a deeply personal experience. When a close friend’s four-year-old daughter suffered a serious personal injury, her family spent two difficult years trying to find an attorney. They asked friends, searched online, and contacted law firms and bar referrals, only to be told “not a fit” more than 140 times. During that time, the family faced mounting medical bills, missed work, and ongoing uncertainty. Geddis, then a senior executive at Allganize.AI which provides Agentic-AI for Fortune 500s, purchased millions of court records and spun up an AI / ML model over a weekend, and uncovered a skilled trial attorney who had been overlooked. One demand letter later, the child’s life was forever changed.

That experience shaped both the company’s mission and its structure. ReferU.AI is a Public Benefit Corporation and funds the ReferU.AI Foundation, a 501(c)(3) that supports low- to-no-cost, experienced legal representation for children whose voices are too often unheard. Equal access to justice begins with equal access to information and representation. At Business Honor, we had the privilege of interviewing Joel Geddis, Founder & CEO of ReferU.AI, to explore how his AI-powered platform is reshaping access to legal representation. Improving outcomes through data-based matching, and advancing digital transformation across the legal sector.

What are the core services that ReferU.AI provides, and how have those offerings evolved over time?

At ReferU.AI, we like to call ourselves the “Tinder for Finding an Attorney.” Users describe their legal needs in plain language to our AI intake agent, “Link,” and we use AI and machine learning to analyze billions of court records to match them with attorneys who have demonstrable experience in similar cases. Once the match is made, we even book the consultation. The platform is free for consumers.

While our core offering hasn’t changed, our approach has. Early on, we relied on docket-level data, the court clerk’s official index of activity. Although machine-readable and has less latency, it’s coarse. Two cases with the same label can differ dramatically in facts and outcome difficulty. For example, in State of Texas v. Benjamin Elliott, a 17-year-old boy tragically killed his twin sister, allegedly, during a sleepwalking episode. His strongest defense, parasomnia, had succeeded in only a handful of cases before. Despite facing a potential 99-year sentence and a 40-year plea deal offered by the prosecution, his defense attorney, Wes Rucker, famously secured a 15-year sentence. While docket-level data would label this case a “loss,” based on the context, many would consider it a significant legal success.

We realized win-rates alone are misleading, so we built a two-stage pipeline: machine learning plus attorney, venue, and judicial analytics on dockets followed by retrieval-augmented generation to find the nuance in filings, arguments, and opinions. It cost more and delayed our launch, but at ReferU.AI, we choose accuracy and better outcomes over convenience.

How does ReferU.AI’s evidence-based matching system differ from traditional lawyer directories and search engines?

At ReferU.AI, we do things differently. Unlike search engines and traditional lawyer directories, which match clients to attorneys based on marketing signals rather than evidence, we prohibit attorney advertising and keep all external influence out of our matchmaking process. Search engine optimization rewards long-form guides, keyword density, backlinks, and technical performance, while paid advertising prioritizes budget, bidding strategy, and ad quality. Traditional lawyer directories often rely on location, self-reported practice areas, or limited profile information, with visibility improved by subscription fees or paid features. Even bar-operated lawyer referral services, which aim to reduce bias, are often required to assign clients largely at random, essentially flipping through a digital Rolodex.

At ReferU.AI, we use a case-similarity index to match users with attorneys who have demonstrable experience in similar fact patterns. Link, our AI-powered legal intake assistant, further improves matches by collecting user-specific preferences like budget, payment options, language, and accessibility requirements. Assuming the attorney has made this information publicly available, we incorporate it into the search to ensure the match fits both the case and the client’s unique needs.

For me, trust is everything. Our AI-powered platform ensures matches are based on evidence of experience, providing fairness, accuracy, and integrity in every connection.

What challenges did you face in building a globally distributed, fully remote team, and how has that impacted product development?

I’ve worked in every environment, full-time on-site, hybrid, and fully remote. Each has its own advantages and challenges, but my decision to hold a fully distributed team came from what I learned as a graduate student in strategic human resource management and as a former technical recruiter.

In the beginning, a remote model allows us to attract top talent by casting the widest net. Many tech professionals prefer fully remote work, giving us a competitive edge in recruiting. Second, it improves retention because team members can achieve better work-life balance and live without geographic constraints. Finally, by working across time zones, we can accelerate progress in sprints, and there’s always a “handoff” between regions so work continues around the clock.

The main challenge is communication. If a blocker isn’t clearly articulated in the handoff, the next time zone can wait up to twelve hours for clarification, which slows momentum. To address this, we’ve become document-first: every handoff, question, and update is clearly articulated in writing so the next team can move forward immediately, improving both clarity and efficiency.

How does ReferU.AI support law firms in receiving more qualified consultations while reducing “not-a-fit” inquiries?

The core challenge in legal matching is “fit.” Studies show that 77% of attorneys don’t respond to consultation emails and 48% of calls. It’s not due to indifference, but because the case isn’t the right fit for their practice. Experience matters most: 43% of consumers and 50% of attorneys say it’s the single most important factor when deciding who to work with.

At ReferU.AI, our matchmaking process solves this by transparently communicating attorney experience to clients. By mining court records for similar cases, we ensure attorneys receive inquiries aligned with their experience and preferred practice areas, while clients can feel confident while connecting with someone who truly understands their case.

Looking forward, what innovations or expansions do you see shaping the future of ReferU.AI and legal tech as a whole?

We have an ambitious roadmap with quarterly launches through 2026, focused on deepening the evidence, tightening workflows, and building trust without adding friction. Legal tech is moving fast, and while predictions have a short shelf life, the winners will be tools that change from assistive to agentic and earn user trust.

Joel Geddis | Founder and CEO

"ReferU.AI uses case-similarity analysis, drawing on public court records, board certifications, and bar data, to identify attorneys with substantial experience in similar cases, giving clients a stronger starting point when deciding who to contact."