He Got Into Insurance for Revenge. Then He Built the Library the Industry Never Had.

Daniel Katz studied insurance for one reason. Revenge.

His family had a claim denied at a moment when they needed it most. Every lawyer he approached wouldn’t touch the case. By the time he was old enough to actually fight it, the statute of limitations had expired. So instead, he decided to learn the industry well enough to understand exactly how it had failed his family, and to make sure it couldn’t happen to anyone else the same way.

He never got his revenge in the courtroom. What he got instead, over a decade working at Willis, an MGA, and as a fraud investigator, was a much clearer view of a different problem: insurance is an industry that moves forward by looking backward, and yet almost nobody actually uses the backward-looking data they already have.

That observation became RiskRemedy. In Episode 172 of InsurTechTalk, Daniel walked me through what book intelligence actually means, how a broker used his platform to grow from 20 to 37 employees while raising entry-level pay, and why he’d rather hand out $100,000 worth of caviar on a New York City sidewalk than pay for a conference booth.

About Daniel Katz

Daniel Katz is the Founder of RiskRemedy, a book intelligence platform built for brokers, agents, and insurance consultants. He began studying risk management and insurance at St. John’s University in 2013 and entered the insurance workforce in 2017, working across real estate, hospitality, and construction lines at Willis. He went on to work at an MGA, investigate insurance fraud on the MGA’s behalf, and manage a compliance team of roughly ten people in affordable housing risk before leaving to found RiskRemedy nearly two years ago. He is also a longtime practitioner of traditional Shotokan Karate and runs a separate caviar importing business, which later became an unlikely marketing tool for RiskRemedy itself.

What RiskRemedy Actually Does

RiskRemedy is built around a simple but underused idea: every policy review, every endorsement comment, every negotiation a broker or consultant has ever done represents institutional knowledge that almost never gets captured, let alone reused. That knowledge typically lives in one person’s head, and it disappears the moment that person retires, gets sick, takes vacation, or leaves the company.

The Insurance Library

Daniel’s own reference point is the insurance library at St. John’s University’s Manhattan campus, reportedly the largest of its kind. RiskRemedy is built to function the same way, except instead of static books, it holds a company’s own placements, policies, and the commentary generated around them, and makes all of it queryable.

• Approaching two million endorsements reviewed, each with commentary on why an endorsement is favorable or problematic
• Coverage spans every state, since insurance regulation and language vary state to state
• Coverage spans multiple industries, since the same endorsement can be a serious problem in one vertical and a non-issue in another
• The platform is explicitly positioned as more than a basic AI chat layer over documents. It is described as a librarian with access to a company’s full historical record, not just a single conversation’s context window

Lines of Business

RiskRemedy focuses roughly 90 percent on commercial P&C, with the remaining 10 percent in personal lines, primarily for high-net-worth individuals. The strongest verticals in the data set are real estate, construction, and hospitality.

Who Actually Buys It

RiskRemedy’s go-to-market strategy deliberately avoided the obvious path.

Starting With Consultants, Not Brokers

Rather than chasing the largest brokerages first, Daniel started with independent insurance consultants, a market of thousands of small operators with no dedicated technology built for them despite holding some of the most rigorous, detail-level coverage expertise in the industry. It took RiskRemedy roughly 13 months to build technology precise enough to meet that bar.

The payoff: consultants using RiskRemedy have grown from typical industry benchmarks of around $1 million in revenue per employee to approximately $2.5 million per employee, driven by compounding knowledge that used to require rereading the same endorsement from scratch every time.

The Client Roster Today

• One of the top 10 brokers in the country
• Two of the top 5 brokers
• Three of the top 25 overall
• A current team of three at RiskRemedy, with a fourth and fifth hire in progress

The Broker Growth Story

One early client, a brokerage with about 20 employees, originally asked RiskRemedy for policy comparison technology and planned to hire six additional back-office staff to manage the workload. Four months later, Daniel checked in expecting to hear about the new hires. The broker had never made them.

Instead, using the capital he had budgeted for those hires, the broker added four placement brokers, two producers, and opened an entirely new claims team, expanding the business rather than just processing more paperwork with more headcount. Entry-level salary at the firm rose from $42,000 to $75,000. As of the recording, the brokerage had grown from 20 to 37 employees in under two years.

Construction Insurance: How Endorsements Spread State to State

This is one of the more genuinely novel data insights from the conversation.

New York and California function as the earliest adopters of new insurance endorsements, particularly for first-party and third-party coverage. Florida has emerged as a third influential state, primarily on property coverage. Once an endorsement appears in one of these states, RiskRemedy’s data shows it beginning to spread into adjacent states, New Jersey, Pennsylvania, Connecticut for the New York and California effect, and states north of Florida for property-specific coverage.

• Fully exclusionary endorsements are climbing roughly 5 percent in these adjacent states
• More conditional endorsements, described as a “soft hammer” rather than a hard exclusion, are climbing closer to 11 percent
• This trend is visible even as the broader market is softening, and it is showing up in RiskRemedy’s data before major brokerages are able to publish it in their own market reports

The practical value for a broker: instead of pitching a client with a generic news story about a similar loss elsewhere, a broker can show a client a specific, quantified increase in the likelihood a particular exclusion is coming to their policy.

The $100,000 Caviar Booth

Daniel’s approach to conference marketing at Insure Tech Insights in New York is worth recounting in detail, because it is a genuinely unusual growth story.

Rather than pay roughly $50,000 for a traditional expo booth, competing for attention against dozens of similar setups, Daniel set up a canopy on the public sidewalk near the venue’s rideshare pickup area and gave away $100,000 worth of caviar at market price, sourced through his own importing business. He researched New York City sidewalk vending law in advance to do it legally.

The result: roughly 250 people stopped at peak hour, including CIOs, COOs, and CEOs of major carriers Daniel says he would not otherwise have had access to. Conversations regularly ran 15 to 20 minutes, well beyond what a typical booth interaction allows, in part because Daniel’s opening question was never about RiskRemedy but about what the other person actually wanted out of their own work. A keynote speaker at the event stopped by and was reportedly impressed enough to ask who had organized it. The answer was one person, Daniel, with no team and no outside funding for the stunt.

The AI and Notion Stack Behind a Three-Person Company

Daniel is candid about how heavily he personally relies on Claude to operate at a scale that would typically require a much larger team.

• He describes his own target as reaching the productivity of a much larger competitor, framed the same way he encourages brokerages to think about their own employees: if the industry benchmark is a million dollars of revenue per employee, how do you get your own people, or yourself, to a million and a half
• Claude and Notion together form a running relationship database. Daniel keeps detailed personal context on the people he meets, from a contact’s dog’s name to a family member’s birthday, so that every follow-up conversation is genuinely personal rather than a generic pitch
• This same relationship-context gap, the difference between an AMS as a system of record and the messy, many-to-many reality of how brokers, underwriters, and clients actually interact, is the specific problem Daniel and Gilad discussed in relation to Backstory, a relationship intelligence platform and a current InsurTechTalk partner

The Insurance Talent Crisis, By the Numbers

Daniel surveyed roughly 1,500 insurance professionals across all levels of seniority in 2025, and the generational gap he found is stark.

• Among professionals over 40, roughly 91 percent read at least one insurance policy cover to cover every year, with some doing so as many as 15 times annually
• Among professionals under 40, fewer than 6 percent had ever read a full policy cover to cover
• A CIO at a top-15 insurance company told Daniel directly that his organization cares far more about rate than about policy verbiage, since verbiage is open to interpretation while rate is a clean, comparable number

Daniel’s concern is structural, not nostalgic. Insurance functions on essentially one financial mechanism: taking in money through underwriting and paying it out through claims, all of it governed by contract language. If the people underwriting and brokering that contract increasingly don’t read it, Daniel believes the industry risks a genuine, fundamental loss of public faith in insurance as an institution, not just a talent pipeline problem.

He connects this directly to a McKinsey finding he cited: someone who left the insurance industry in 2006 and returned 20 years later would find the day-to-day work largely unchanged and could reintegrate easily. Meanwhile, operations budgets at major brokerages are growing at what Daniel describes as an exponential rate relative to per-employee cost, a sign that simply adding headcount has not solved the underlying knowledge transfer problem.

Key Takeaways

• Book intelligence means treating a company’s entire historical record of policies, endorsements, and internal commentary as a queryable library, not a filing cabinet
• Starting with underserved insurance consultants rather than large brokers built RiskRemedy a uniquely rigorous data set before it ever reached bigger clients
• Endorsement trends in New York, California, and increasingly Florida reliably predict which coverage exclusions are about to spread to neighboring states
• Freeing up broker capacity through technology can fund new hires and higher entry-level salaries rather than simply cutting headcount
• A memorable, unconventional marketing stunt can outperform a traditional conference booth by orders of magnitude in access and conversation quality
• The generational gap in policy literacy, 91 percent of professionals over 40 versus under 6 percent under 40, is a genuine structural risk to the industry, not just a talent pipeline statistic