ReferU.AI is an AI-powered platform that analyzes billions of court records and legal analytics to match individuals with attorneys who have verified, specific experience in cases similar to their own. Pulse 2.0 interviewed ReferU.AI founder and CEO Joel Geddis to learn more.
Joel Geddis’ Background
Could you tell me more about your background? Geddis said:
“My background is a bit squiggly because I value breadth and learning from new challenges. It spans three industries – media and entertainment, brick and mortar retail, and technology – including three startups, a family-owned business, and a Fortune 50. But, what’s prepared me most as a solopreneur is that, by now, I’ve served in a variety of roles such as creative and marketing, sales and operations, logistics, security and compliance, and human resources. Most recently, I was Chief of Staff at Allganize.AI, a leading provider of Agentic-AI for Fortune 500s, where I ran all business operations for North America.”
Formation Of The Company
How did the idea for the company come together? Geddis shared:
“ReferU.AI was born from a personal experience: a close friend’s four-year-old daughter suffered a traumatic injury. While they were under the weight of one of life’s heaviest moments, the process of finding an attorney wasn’t an easy lift. After repeating their story to 140 law firms over 2 years, the unanimous verdict was ‘not a fit for our practice.’”
“I thought, ‘dating apps figured out chemistry with AI & ML, so why do we still search for attorneys like it’s 1995?’”
“So, I went to that courthouse, purchased millions of court records and, over a weekend, spun-up an AI/ML agent to search the dockets and filings for cases with similar fact patterns. The question I was trying to answer was, ‘Who’s winning these types of cases?’ The agent returned a young, previously overlooked, trial attorney, and their 141st call was ‘Ah, heck yeah. I’m all in!’”
“He wrote a single demand letter and the young girl received a substantial settlement.”
“That agent became our MVP. We’re solving three problems:
First, only 36% of clients report the intake process as seamless. By automating research, consideration, and contact, users don’t have to tell their story on repeat.
Second, 50% of consumers and 43% of attorneys believe that experience in similar matters is the most important factor when determining ‘fit.’ By surfacing attorneys with prior related experience, users and attorneys can feel confident in their decision.
Finally, 72% of attorneys reported needing higher-quality, not more, leads. By mining court records and, when possible, language, accessibility, and payment options, “not a fits” are naturally filtered out.”
“Importantly, we don’t monetize our platform through attorney subscriptions or advertising, nor do we allow any outside influence over our algorithm. And, contrary to some lawyer directories, which generally have up to ~400,000 attorney profiles (those who choose to pay a monthly subscription), all about 1.3M active and licensed attorneys are available on our platform. Common sense says that more options often lead to more just outcomes.”
“We’re also structured for impact. Sadly, one study revealed that only 23% of children in a complex child-welfare system designed to protect them were represented by attorneys and, in 22% of cases, the child was never even mentioned in court. The implications are profound and long-lasting: unfair hearings, unjust removals, unsafe placements, lower educational achievement, and higher likelihood of homelessness and incarceration. That research aligns with our origin story: the smallest voices in court need experienced counsel to speak up for them.”
“As a Public Benefit Corporation, we pledge a portion of profits to our non-profit, The ReferU.AI Foundation, a 501(c)(3) dedicated to supporting child victims of abuse and neglect. We operationalize social impact by providing a fund for attorneys and vetted child-welfare organizations to withdraw from in order to provide no-cost dedicated counsel to children in court.”
Favorite Memories
What are your favorite memories so far at ReferU.AI? Geddis reflected:
“Every time Link, our AI Intake Agent, matches someone with an experienced attorney, we’re reminded that AI isn’t just for enterprise automation. Instead, we see a direct link between the code we ship and the impact on each individual users’ lives.”
Core Products
What are the company’s core products and features? Geddis explained:
“We’re the ‘Tinder for Finding an Attorney.’ Users visit our platform, ReferU.AI, and describe their legal needs in plain language. Our AI-Agent, ‘Link,’ interprets their needs, mines millions of court records and documents, and instantly matches them with attorneys who have evidence of success in similar fact patterns. It’s truly user-first: it’s free to use, returns unbiased matches, and even considers language preferences, budgets, payment options, and accessibility needs. Then, we automate attorney outreach and scheduling.”
Challenges Faced
Have you faced any challenges in your sector recently and how did you overcome them? Geddis acknowledged:
“Yes, court data are fragmented. There are thousands of courts across the U.S., and access to records varies by each venue. Are they even digitized? How’s it delivered? APIs, FTP, hard copies, or a mailed-in hard drive with a return label? And, how’s it structured and labeled? Terminology isn’t standardized; data entry differs from court to court.”
“We’ve invested heavily in coverage, normalization, and quality assurance — building ingestion pipelines across heterogeneous sources; mapping codes, aliases, and abbreviations; and continuously validating with human-in-the-loop review. It’s unglamorous work, but it’s essential to deliver precise matches.”
“And, another thing: privacy and security. While consumers are growing increasingly comfortable sharing the legal and medical questions with Google, ChatGPT, Claude, and the like, we’re a relatively new player in the game. That’s why privacy and security is baked into both our business model and product architecture.”
“In regards to our business model, when someone interacts with our AI-Agent, “Link,” they’re speaking with an authorized agent of ReferU.AI Law, PLLC, a D.C.-based law firm. As such, intake sessions are protected under attorney-client privilege and prospective-client confidentiality.”
“That matters.”
“In one example, G.R. v. Doe 1, Los Angeles Superior Court, No. 22STCV38612, defendants in a civil childhood-sexual-abuse case subpoenaed a third-party attorney lead-generation website for communications and documents received from or for the plaintiff. The court ordered the third party to turn over records.”
“In another, a San Francisco criminal matter, the District Attorney subpoenaed the Bar Association of San Francisco’s Lawyer Referral and Information Service. The subpoena sought what a criminal defendant may have said to BASF staff while seeking a referral.”
“And, why does this matter?”
As a law firm, we have an ethical obligation to keep intake sessions confidential and attorney-client privilege protects prospective client communications from forced disclosures.*”
“In regards to our product, we treat privacy and security as a core architectural design constraint, not like bolted-on scaffolding as an afterthought. If a feature can’t be deployed safely, it doesn’t make it on the roadmap.”
“We don’t train our large language models on user conversations and we redact personally identifiable information pre-processing. Data is encrypted in transit and at rest, we have strict data minimization and retention controls. We hire ethical hackers monthly who attempt to breach our advanced security measures and make those results available upon request. We also offer bug bounty rewards up to $20,000.”
“Promptly after a chat session ends, session content is removed from active chat processing systems, and the necessary user information is securely transmitted to and temporarily stored in Salesforce CRM. In Salesforce, PII and conversations are encrypted with tamper-proof hashing so that no one in our organization, not even I, can see it.”
“Salesforce publicly lists independent SOC 1, SOC 2, and SOC 3 audits, along with ISO/IEC 27001, ISO/IEC 27017, and ISO/IEC 27018 security certifications, and we’re currently undergoing our SOC 2 Type II audit.”
Evolution Of The Company’s Technology
How has the company’s technology evolved since launching? Geddis noted:
“We began with docket-level data, which are the court clerk’s official entries of case activity. It’s useful, lower-compute costs and latency, but coarse. Early on, we recognized that crude signals like ‘win rate’ can be misleading because they’re easily influenced by what cases attorneys choose to take. And, while two cases may share a label, they might diverge materially in fact patterns and outcome-difficulty. “Thus, ‘what’s a win?” isn’t uniform.’
Today, we use a two-stage pipeline: machine learning on docket-level data for a rough search and then Agentic-RAG (retrieval-augmented generation) over billions of primary documents like complaints, motions, orders, and transcripts, plus robust attorney, opposing counsel, and judicial analytics. Then, we compute a case-similarity index between a user’s needs and the attorney’s prior experience in similar fact patterns.”
“Doing it this way significantly delayed our launch and, while it’s not perfect, we believe it’s better than the status quo and we’re constantly fine-tuning to deliver on our mission of Smarter Results. Better Representation. The Best Outcomes.”
Significant Milestones
What have been some of the company’s most significant milestones? Geddis cited:
“On day-one of go-live, our automated matchmaking service helped our first customer.”
Customer Success Stories
Can you share any specific customer success stories? Geddis highlighted:
“Privacy is paramount, but these users have given us permission to share anonymous success stories.”
“In one matter, a user was injured by factory equipment lost in a food item. ReferU.AI matched them with an experienced food and product liability attorney. Their claim accounted for medical expenses, lost income, pain and suffering, and punitive exposure.”
“Another was an assault against a public official during a mental break. Facing years of confinement. ReferU.AI matched them with two criminal defense attorneys who represented a near-identical case in that venue. They secured a treatment plan in lieu of incarceration. By getting them the help they needed, their outlook on life changed dramatically.”
“In a civil wrongful death action, the estate’s attorney withdrew when they couldn’t find the right expert witness. After several months of searching, ReferU.AI connected them with an experienced wrongful-death attorney and the expert witness they needed, allowing the estate to successfully bring the suit forward.”
Funding/Revenue
Are you able to discuss funding and/or revenue metrics? Geddis revealed:
“We’re post-revenue with a strong sales pipeline and long runway. We’re bootstrapped by design and, while inbound investment and acquisition opportunities are real, we’re more interested in strategic advisors and partners than cash.”
Total Addressable Market (TAM)
What total addressable market (TAM) size is the company pursuing? Geddis assessed:
“In the U.S., legal services represent a market of about $396 billion.”
Differentiation From The Competition
What differentiates the company from its competition? Geddis affirmed:
“Traditional searches for attorneys — search engine optimization and ads, billboards, lawyer directories, lead-gen marketplaces, etc. — reward advertising budgets; marketing hearsay over evidence of experience.”
“Lawyer directories and marketplaces are generally subscription, pay-to-rank, and pay-per-lead models. They usually rely on archaic regex keyword- and filter-based searches and opaque ‘best match’ logic. But, that ‘logic’ are the types of cases the attorneys want, not necessarily where they’re experienced.”
“While bar-operated lawyer referral services provide a public benefit and aren’t biased towards attorney advertising, their vetting process is surface level: licensure, disciplinary record, bar standing, and preferred case types. And most state bars require lawyer referral services to assign attorneys by random allotment. Think, ‘blindly flipping through a digital rolodex.’”
“ReferU.AI is matchmaking, not marketing. Every licensed practicing attorney is accessible on our platform — there’s no ‘claim your profile,’ subscription fee, paid rankings, or any other type of attorney advertising that influences the other commercial solutions.”
Future Company Goals
What are the company’s future goals? Geddis emphasized:
“In the near term, our focus is shipping three core products, hardening them with real-world usage, and proving repeatable impact. Our north-star metric is making our first donation to The ReferU.AI Foundation, completing the link between why we started and how we improve access to justice. Beyond that, we have a sequenced roadmap through 2027 with additional products that extend our mission.”
Team Composition
What’s your team composition, and is there anyone you’d like to recognize? Geddos described:
“We’re a young, diverse, and globally distributed team of 20. We consist of AI/ML engineers, data scientists, PhD-level AI researchers, UI/UX designers, business analysts, go-to-market catalysts, and product/program managers. Their backgrounds range from startup war rooms, top-tier technical research universities, and Fortune 500s.”
“Anyone I’d like to recognize? That’s a long list, so I’ll shout out the one person I work most closely with: Diksha Nasa, Senior Product Manager. I’m a strategic, creative thinker who can over-index on vision; Diksha provides the sanity checks, structure, and execution rigor that translate innovative ideas into shipped products and real outcomes for users.”
Additional Thoughts
Any closing thoughts? Geddis concluded:
“I want to thank you for the opportunity to interview with Pulse 2.0 and for bringing awareness to what we’re trying to accomplish at ReferU.AI.”
“When someone needs an attorney, generally, they’re facing one of life’s formidable challenges at a time when every step feels expensive and uncertain: a divorce, personal injury, deportation, or even their freedom. And, some decisions should be transparent and based on evidence, not slogans.”
“ReferU.AI turns raw court data into actionable insights so people can find their advocate based on court record evidence instead of marketing hearsay. That’s why we exist: to deliver Smarter Results, Better Representation, and the Best Outcomes.”
* Rules may vary by jurisdiction.
