Advertise on podcast: InsurTechTalk

A few episodes ago, Cheri Martinen revealed that GL policies are now excluding AI-generated content. The question she left unanswered: who verifies whether content is AI-generated when a claim arises? Yagub Rahimov, Founder & CEO of Polygraf.ai, is one of the answers. In Episode 165 of InsurTechTalk, Yagub explains how his AI behavioral control platform addresses the three AI threats insurance companies are not ready for: deepfake claims fraud, shadow AI data leakage, and the context gap that checkbox-based claims processing can never close. In this episode: - Why 96% of cyber failures are behavioral, not technical — and why most security tools are built for the other 4% - The Beverly Hills problem: why regex-based DLP misses 32% of email-related data leakage and far more for contextual sensitive data - Why Polygraf runs entirely locally — no API exposure, no cloud, air-gapped when required - How deepfake claims fraud is evolving from isolated documents to coordinated fraud rings - The shadow AI crisis in insurance: what a CISO's "signed letter" solution reveals about how carriers are actually governing AI today - Why the biggest gap in insurance claims processing is context — and how AI can close it - Yagub's own Airbnb scam experience and what it reveals about embedded insurance failures 🔗 Learn more: https://www.polygraf.ai/ 🔗 Connect with Yagub: https://www.linkedin.com/in/yrahimov/ InsurTechTalk is the podcast for insurance professionals, InsurTech founders, and anyone who wants a no-BS look at where the industry is headed. 🔔 Subscribe for new episodes every week.

In 1998, Jon Kelly co-founded eCoverage — backed by SoftBank and Accel — with a one-line mandate: build the online version of Geico. Eleven months later, with no cloud infrastructure and no playbook, they issued their first policy. It became the first startup to sell car insurance online. In Episode 164 of InsurTechTalk, Jon Kelly — now CEO of Modern Metric — walks through three decades at the frontier of insurance technology: the dot-com collapse that killed his Series C overnight, inventing the click wall that became a billion-dollar lead generation category, and the six-week nightmare of buying his own Chubb policy that led him to build the private client insurance platform he needed. In this episode: 1. Building a full-stack online carrier in 11 months in 1998 — with no AWS, no cloud, no precedent 2. What happened when the Barron's article hit and the funding "light switch" turned off overnight 3. The accidental origin of the click wall and why lead generation has such a complicated reputation 4. Why six weeks to get a Chubb policy became the founding thesis for Kelly Klee 5. Why direct-to-consumer carriers built great technology decades before independent brokerages could 6. The sharp line between where AI helps (data ingestion, complex analysis) and where it does not scale (team-wide workflow)6. Modern Metric's Discover platform and why standardizing process matters more than AI tools 7. His closing argument: the industry needs operational excellence more than it needs another AI feature 🔗 Learn more: https://www.modernmetric.com/ InsurTechTalk is the podcast for insurance professionals, InsurTech founders, and anyone who wants a no-BS look at where the industry is headed. 🔔 Subscribe for new episodes every week.

Jonathan Crystal opened Crystal Venture Partners on November 1, 2022. ChatGPT launched 30 days later. That timing tells you everything about where this conversation goes.In Episode 163 of InsurTechTalk, the third-generation insurance operator, former CFO of Crystal & Company, and Founder of Crystal Venture Partners joins me for a wide-ranging conversation on what he actually looks for in a founder, why he does not believe in InsurTech anymore, and what the industry is completely missing in its obsession with AI.In this episode:- Why Crystal Venture Partners is not an InsurTech fund — and why that distinction matters- The three things that make a business durable in an AI world: regulatory navigation, capital efficiency, and trusted relationships- How he backed Will Johnson at Gyde after eight to nine months of conversations before ever seeing a pitch deck- Why he skips the first six slides of every founder's deck — and what he actually wants to know instead- The death of pure SaaS: why outcome-based and gain-share models are replacing seat-based licensing- What venture capital actually means for a founder — and why it is the most expensive capital imaginable- How $150M+ in follow-on capital across 11 portfolio companies in six months validates the thesis- His answer to the closing question: the industry is talking too much about technology and not enough about people🔗 Learn more: https://www.crystalventurepartners.com/🔗 Connect with Jonathan: https://www.linkedin.com/in/jonathancrystal/InsurTechTalk is the podcast for insurance professionals, InsurTech founders, and anyone who wants a no-BS look at where the industry is headed.🔔 Subscribe for new episodes every week.

Everyone talks about the AI boom in terms of compute, models, and software. Justin Levine spends his days thinking about something different: who insures the actual construction of the data centers, energy assets, and projects that make all of it possible.The AI race moved from the cloud to the construction site. And the construction site has an insurance problem. In Episode 162 of InsurTechTalk, Shepherd Co-Founder & CEO Justin Levine — civil engineer turned construction risk manager turned InsurTech founder — explains how his company is building the first behavior-based commercial underwriting platform, why the Coalition-Allianz deal is a seminal moment for the MGU model, and how Shepherd Savings is rewarding contractors for technology they already use with up to 25% lower premiums. In this episode: - Why the construction insurance problem is speed and data, not capacity — and why that distinction matters- How Shepherd delivers underwriting indications in minutes by integrating with Procore, Autodesk, DroneDeploy, and OpenSpace- The personal auto telematics parallel: why commercial construction is 20 years behind on behavior-based pricing — and what happens when it catches up- Shepherd Savings: how rewarding existing technology adoption solves the change management problem that killed wearables in workers comp- Why data centers are actually a minority of Shepherd's $400B+ insured portfolio- The seven archetypes of data center construction — and why one size fits none of them- The Coalition-Allianz deal: the first proof point that technology-native MGUs can outperform legacy carrier underwriting- Closed-loop data center developments — power, compute, water, storage — and the new complexity they create for construction insurers- What is next for Shepherd: new programs, new verticals, and announcements expected in Q3 🔗 Learn more about Shepherd: https://www.shepherdinsurance.com/🔗 Connect with Justin: https://www.linkedin.com/in/justindlevine29/ InsurTechTalk is the podcast for insurance professionals, InsurTech founders, and anyone who wants a no-BS look at where the industry is headed. 🔔 Subscribe for new episodes every week.

Cheri Martinen never planned to sell her family's insurance agency. Then she took a meeting that surprised her.In Episode 161 of InsurTechTalk, the fourth-generation agent and Managing Principal at OneDigital shares what made the difference — and drops one of the most practically urgent insights of the season: standard GL policies are now excluding AI-generated content, effective April 2026. Most business owners do not know. Many agents do not know either.In this episode:- Why she sold — and why it was a win for her employees, customers, and family, in that order- What made OneDigital different from every other approach she had received over the years- How to run an agency that sails through due diligence (hint: it is the same as running a good agency)- Why wildland firefighter contractors and water utilities became her niches — and why niche is a moat, not a limitation- The AMS filing cabinet problem that InsurTech still has not solved- Telematics, wildfire prevention, and the agent's role in risk education- The AI content liability gap: GL exclusions, cyber coverage gaps, and the new MGAs launching AI liability products to fill the void🔗 Connect with Cheri: https://www.cherimartinen.com/🔗 Learn more about OneDigital: https://www.onedigital.com/InsurTechTalk is the podcast for insurance professionals, InsurTech founders, and anyone who wants a no-BS look at where the industry is headed.🔔 Subscribe for new episodes every week.

What if AI's role in insurance claims isn't to replace the adjuster — but to be a second set of eyes that never gets tired?Heather Wilson, CEO of CLARA Analytics, joins InsurTechTalk episode 159 to explain how her company's claims AI platform — trained on roughly 10 million claims over a decade — surrounds every workers comp, auto liability, and general liability case 24/7, looking for anything that could take it off course. The result: earlier detection of severity changes, better provider matching, litigation risk flags, and millions in reserving corrections that adjusters often miss simply because cases sit untouched for 90 days at a time.In this episode:• How CLARA's AI models learn from adjuster notes and documents — not just structured data fields (which are often filled out poorly)• The real story of a "finger cut" that escalated to an amputation, and how AI caught the severity shift before a human did• Why CLARA frames its AI as "augmented intelligence" — a safety net, not a replacement — and how that framing drives adoption• The change management playbook for getting senior adjusters to trust (and act on) AI alerts• How CLARA works across the entire insurance ecosystem — carriers from tier one to tier five, TPAs, self-insured corporates, and brokers who white-label CLARA's platform• Why "data orchestration" — not data ownership — is CLARA's real competitive moat• The hard truth about selling claims improvement: "we'll make you more profitable without increasing revenue" is one of the toughest pitches in insurance• Multi-tenant data contribution: how anonymized, aggregated industry data benefits every carrier on the platformHeather's vision for the future: connecting claims intelligence directly to actuarial and policy intelligence — closing the loop between the "left hand and right hand" of insuranceA grounded, practical look at what AI actually does inside a claims shop today — not hype, not science fiction, just better outcomes for claimants, carriers, and employers.🔗 Learn more about CLARA Analytics: https://www.claraanalytics.com/InsurTechTalk is the podcast for insurance professionals, InsurTech founders, and anyone who wants a no-BS look at where the industry is headed.🔔 Subscribe for new episodes every week.

AI is everywhere in insurance right now — but is the industry actually ready for it?In episode 157 of InsurTech Talk, Gilad Shai sits down with Jack Siney, co-founder and CRO of FrontRace, to discuss the gap between AI hype and operational reality.Jack explains why most insurance organizations still struggle with disconnected systems, weak data infrastructure, outdated workflows, and incomplete sales processes — and why simply “adding AI” will not solve those problems.The conversation explores:• Why today’s AI adoption resembles the AOL era of the internet• The hidden operational inefficiencies inside insurance organizations• Why most sales and distribution processes are poorly documented• The difference between automation and true operational intelligence• How AI can uncover why top producers outperform everyone else• Why insurers should focus on data readiness before buying more AI tools• The future of AI-driven sales management and underwriting workflowsA practical conversation for insurers, agencies, MGAs, founders, operators, and investors trying to separate real AI value from noise.Subscribe for more conversations with founders, carriers, brokers, investors, and operators shaping the future of insurance and insurtech.#InsurTech #Insurance #AI #ArtificialIntelligence #SalesTech #DigitalTransformation #InsuranceTechnology #EnterpriseAI #Innovation #Underwriting

What if insurance could pay out automatically — no adjuster, no claim dispute, no waiting? Sid Jha, CEO of Arbol, left a career trading commodities at Citadel to build exactly that. In Episode 158 of InsurTechTalk, Sid breaks down how Arbol went from a blockchain white paper in 2018 to a vertically integrated insurance platform spanning MGAs, a property carrier, and reinsurance — all purpose-built for climate-sensitive risks that legacy carriers are quietly retreating from.In this episode:- Parametric vs. traditional insurance — and why the old "basis risk" critique is stale- How Arbol underwrites climate risk using satellite data, AI, and quant methods- Why parametric is a bridge to alternative capital (hedge funds, tokenized funds, and beyond)- The vertical integration journey: MGA → MGU → carrier acquisition (Centuri, 2024)- Tokenized insurance risk — the original white paper vision that's finally becoming real- How Arbol approaches distribution in a market most brokers didn't know existed- The truth about climate risk: it's far more regionally varied than headlines suggestA rare conversation at the intersection of Wall Street, InsurTech, and climate — with someone who's been building quietly at that crossroads for nearly a decade.🔗 Learn more about Arbol: https://www.arbolmarket.com/🔗 Connect with Sid: https://www.linkedin.com/in/siddhartha-jha-ba4b7534/InsurTechTalk is the podcast for insurance professionals, InsurTech founders, and anyone who wants a no-BS look at where the industry is headed.🔔 Subscribe for new episodes every week.

Surety bonds are one of the least understood—but most critical—segments in the insurance ecosystem.In Episode 156 of InsurTech Talk, I sit down with Gary Eastman, founder of SwiftBonds, to break down how this “unsexy” category quietly supports trillions of dollars in economic activity—and why it’s becoming increasingly relevant again.Gary started as an attorney doing probate work, discovered court bonds, and turned a side project into an 18-year business focused entirely on surety. What followed is a masterclass in niche selection, distribution strategy, and underwriting discipline.We cover:What surety bonds actually are (and why they’re not traditional insurance)The three core categories: contract, court, and commercial bondsHow underwriting works in a “zero-loss assumption” modelWhy distribution—not product—is the real bottleneckThe evolution from referrals → SEO → AI-driven discoveryWhere AI can (and can’t) improve underwriting and operationsWhy specialization beats breadth in building a brokerageThe structural role of bonds in financing, construction, and legal systemsKey insight:Surety isn’t just a product—it’s infrastructure. It guarantees execution, compresses risk timelines, and keeps capital flowing in real-world projects.For founders, investors, and operators in insurance and fintech, this is a category hiding in plain sight.

What happens when your core product works perfectly — but your customers still don't stick around? Adam Price, CEO of Kinetic Comp, joins InsurTechTalk episode 155 to unpack one of the most honest startup pivot stories in InsurTech: how Kinetic went from a wearable device company to a $100M+ workers comp MGA powered by claims oversight software — and why that shift was driven entirely by listening to policyholders and brokers.In this episode: Why wearables work great — but change management kills adoptionThe real reason prevention tech struggles in mid-market workers' compHow Kinetic transitioned from hardware to a software-based claims oversight platformWhat "data fusion" actually means in practice (not just buzzwords)The Nationwide Mutual partnership and what a full-stack carrier relationship really looks likeHow to pitch an MGA to investors — and why the $6 trillion TAM slide won't cut itAI in insurance: Kinetic's honest "5 or 6 out of 10" self-assessmentTips for InsurTech founders building MGAs from scratchWhether you're building in workers comp, launching an MGA, or trying to figure out where prevention tech fits in your product strategy — this one is packed with hard-won lessons.🔗 Learn more about Kinetic: https://kineticcomp.com/

Most insurance leaders focus on distribution, underwriting, or AI. But there’s a quieter constraint slowing everything down: producer compliance and onboarding.In episode 154 of InsurTechTalk, Gilad Shai sits down with Luis Pino, CEO of ProducerFlow, to unpack the infrastructure layer behind insurance distribution—how carriers, MGAs, and agencies actually get producers licensed, appointed, and ready to sell.We go deep on:Why producer onboarding speed directly impacts revenueThe hidden complexity behind licenses, appointments, and compliance workflowsHow poor onboarding creates drop-off before production even startsWhy data quality—not AI—is still the limiting factorThe different risk tolerance between carriers and agencies when adopting AIWhere the real moat is forming: data, workflows, and liability ownershipProducerFlow sits at the intersection of compliance, data, and distribution—turning what’s traditionally a back-office function into a frontline growth driver.If you’re a carrier, MGA, agency, or investor, this is a conversation about the layer of insurance most people ignore—and why that’s where leverage is shifting.

In Episode 153 of InsurtechTalk, Gilad Shai sits down with Will Johnson, CEO and Co-Founder of Gyde, to unpack one of the most overlooked—and misunderstood—segments in insurance: health insurance distribution.This conversation goes beyond the AI hype cycle. It dives into the real bottleneck in insurance: the broker.Will shares how Gyde is building an AI-powered platform designed to help agency owners break through capacity constraints, automate low-value workflows, and return to what actually matters—advising clients and growing their business. From renewal automation to client service and cross-selling, this is a deep look into what a true “broker operating system” looks like in health insurance.We also explore:Why health insurance is fundamentally different from P&C—and harder to fixThe growing pressure on brokers, carriers, and the entire systemHow AI is actually being used (and where it’s not)Why distribution is regaining power after years of disintermediation narrativesThe shift from “selling policies” to becoming a true strategic advisorWill brings a rare perspective shaped by nearly a decade at Oscar Health, where he saw firsthand how critical brokers are—and how underinvested they’ve been.Key takeaway:AI won’t replace brokers. But brokers who use AI will replace those who don’t.

Episode 152 of InsurTechTalk Kasey Roh--Head of US, Upstage AI: Building the “Perfect Marriage” Between OCR and LLMs. What happens when AI stops trying to boil the ocean—and instead solves one of the most painful, expensive problems in enterprise workflows? In this episode, I sit down with Kasey Roh, Head of US at Upstage AI, to unpack how a focused, infrastructure-first approach to AI is quietly transforming industries like insurance, healthcare, and manufacturing. We start somewhere unexpected—yoga, discipline, and managing chaos in the age of AI—and quickly move into one of the most important conversations in enterprise tech today: why most AI implementations fail to scale, and what it actually takes to build systems that compound. Upstage isn’t just another “AI wrapper.” They’ve built both their own OCR engine and a proprietary LLM (SOLAR), designed specifically for document understanding—arguably one of the most overlooked bottlenecks in enterprise operations. Key topics we cover: 1. Why traditional OCR only solves ~10–20% of enterprise document workflows 2. The hidden cost problem of LLMs at scale (tokens, GPUs, inference) 3. Why hallucinations kill enterprise trust—and how to design around them 4. The “OCR + LLM” architecture as a pragmatic middle ground 5. Turning 10 years of unused documents into entirely new insurance products 6. Why AI in enterprises is not about automation—but accountability 7. The real test for AI adoption: does your investment *compound*? We also go deep on: 1. Evaluating AI vendors in a world full of “AI-washing” 2. The importance of owning core models vs. relying on wrappers 3. Human-in-the-loop systems as a feature—not a limitation4. The future of agentic workflows and governance This is a must-watch for operators, investors, and anyone trying to separate signal from noise in the AI hype cycle. Timestamps: 00:00 – Yoga, chaos, and entering the AI world 04:30 – What Upstage actually does (OCR + proprietary LLM) 12:00 – Why enterprise OCR is fundamentally broken 20:10 – The cost vs. accuracy tradeoff in LLMs 28:45 – Unlocking value from “dead” enterprise data 38:20 – AI vs. humans: accountability and trust 47:10 – Evaluating AI vendors (what actually matters) 55:00 – The compounding framework for AI investments About Kasey Roh: Formerly at Tesla and Meta, Kasey transitioned from corporate finance and venture investing into operating, now leading Upstage’s US expansion. Upstage has raised ~$150M and is scaling rapidly across global enterprise markets.

InsurTech Under Fire — Resilience, Insurtech, and the Future of Insurance with Kobi BendelakWhat does resilience actually look like?In this episode, we sit down with Kobi Bendelak, who shares an unbelievable story: leading a delegation of 20 startups and 40 people out of Israel during wartime—when all flights were canceled.The solution?A 40-hour journey through the desert, crossing into Egypt, flying via Athens, and pushing forward to London.Not for optics. For trust.Because in insurance, showing up isn’t optional.What we cover in this episode:* The “reverse exodus” journey from Israel via Egypt to London under fire* Why trust is the foundation of the insurance industry* How the Israeli insurtech ecosystem continues to grow—even during a crisis* The evolution of insurtech: from “tech tourism” to real business outcomes* Why B2C insurtech startups had to rethink distribution (hint: brokers win)* The rise of global collaboration across ecosystems* A bold prediction: AI could replace core insurance systems sooner than expectedChapters:00:00 – Introduction02:10 – The 40-hour journey to London08:45 – Resilience and trust in insurance15:20 – Growth of the Israeli insurtech ecosystem24:30 – From “tech tourism” to real business32:10 – B2C vs brokers: hard lessons41:50 – Global expansion and new markets49:30 – AI and the future of insurance core systemsAbout InsurTech IsraelInsurTech Israel supports over 250 startups through investments, accelerators, global roadshows, and partnerships—making it one of the most active insurtech ecosystems in the world.Key takeaway:Resilience isn’t a story you tell.It’s what you do when everything works against you—and you show up anyway.👍 Like, comment, and subscribe for more conversations with leaders shaping the future of insurance and technology.#Insurtech #AI #Insurance #Startups #Innovation #Leadership

What happens when AI doesn’t just support insurance workflows—but replaces them?In Episode 150 of InsurTechTalk, James Benham breaks down what’s really happening inside insurance companies right now:Why ~30% of the workforce is stuck in manual clerical workThe growing shortage of underwriters and adjustersHow AI is reshaping core systems, workflows, and decision-makingWhy most “AI adoption” today is risky, untested, and non-compliantAnd what it actually takes to build a modern, AI-native insurance stackFrom rebuilding policy and claims systems to embedding AI directly into workflows, this conversation goes beyond hype—and into execution.If you want to understand where insurance is really headed, start here.