"Every industry can benefit from captives, bringing more control and transparency to how insurance is managed," states Illia Kasian, Chief Technology Officer and co-founder of Matcha, the San Francisco insurtech company transforming how businesses approach captive insurance. The statement captures a fundamental shift unfolding across alternative risk markets where artificial intelligence meets centuries-old insurance economics.
Kasian's platform arrives at a moment when captive insurance has reached unprecedented scale. The global market surged to approximately 8,000 captives writing $50 billion in premiums during 2024, according to Risk Management Advisors, while Ernst & Young reports captives now command nearly 25 percent of the overall commercial insurance market. Yet most mid-market companies remain locked out of these arrangements due to operational complexity and resource constraints. Matcha's answer involves deploying agentic AI throughout the entire captive lifecycle, from policy administration through accounting, eliminating barriers that have traditionally required dedicated insurance and IT departments.
Automation Converges with Alternative Risk Transfer
The convergence between autonomous AI systems and captive insurance reflects broader technological momentum across financial services. Insurance companies moved from 8 percent full AI adoption in 2024 to 34 percent in 2025, representing a 325 percent year-over-year increase, according to InsuranceNewsNet analysis. Gartner forecasts that 33 percent of enterprise software applications will incorporate agentic AI by 2028, up from less than 1 percent in 2024.
Kasian's technical approach distinguishes itself through what he describes as a compliance-first AI architecture. Rather than treating machine learning models as opaque decision engines, Matcha embeds strict audit logging, human oversight checkpoints, and traceable decision pathways throughout its platform. "Clean data pipelines, interoperable systems, and AI-driven insights reduce leakage and give stakeholders the timely information they need to manage performance," Kasian explains. The methodology addresses regulatory scrutiny while maintaining the operational velocity that makes captives attractive alternatives to traditional coverage.
The platform handles incident reporting, claims management, accounting, policy administration, and captive reporting through integrated workflows. Large language models process unstructured data from submissions and documents, while autonomous agents execute multi-step tasks like claims management and premium calculations. Mid-market companies can launch captive programs through integrated platforms that coordinate licensed captive managers, actuaries, and administrators without requiring internal specialized teams. Associations and industry groups gain similar capabilities, allowing entire sectors to pool risk collaboratively.
Matcha's early traction validates market appetite for technological solutions to operational friction. The company recently closed a multi-million dollar pre-seed funding round and onboarded its first customers through pilot programs with existing captive service providers. Specific client numbers remain confidential, yet adoption patterns suggest strong demand exists for AI-powered technology that enables captive administrators to manage operations more efficiently among organizations that previously lacked access to these risk management tools.
From Fraud Detection to Insurance Infrastructure
Kasian's trajectory through machine learning applications spans diverse high-stakes domains before converging on insurance technology. His work at Quarks Tech produced fraud detection models that isolated approximately 90 percent of fraudulent activity within the first five minutes of transactions, while image moderation systems increased automated content review from 5 percent to 85 percent. Those improvements translated to roughly $100,000 in annual cost savings through reduced manual review requirements.
The technical foundation supporting such performance gains carried forward when Kasian joined Corgi Insurance as a founding engineer. Corgi earned acceptance into Y Combinator's Summer 2024 batch, where Kasian built core infrastructure and tooling that enabled rapid launch and reliable scaling. He assembled the early technical team and established architectural patterns for handling insurance workflows programmatically. The experience provided direct exposure to how legacy systems and manual processes constrain innovation in property and casualty markets.
Kasian's educational background in applied mathematics from Taras Shevchenko National University of Kyiv emphasized probability theory, statistics, optimization, and machine learning fundamentals. He supplemented formal training with intensive self-directed study in large language models, distributed systems, and financial risk modeling. The combination equips him to navigate both the mathematical complexities of risk quantification and the engineering challenges of building production-grade systems for regulated industries.
His proprietary research spans ML architectures for large-scale content moderation, early fraud detection, and AI-powered back-office operations for insurance carriers. While this work remains unpatented, it has shaped how multiple organizations approach AI deployment in production environments. Kasian co-founded Goodsend between roles, shipping AI-driven products across ML models, backend systems, and mobile applications. The end-to-end product ownership experience informs Matcha's full-stack approach to captive administration technology.
Market Forces Accelerate Captive Formations
The broader captive insurance market shows robust expansion across multiple dimensions. The global captive insurance market reached approximately $50 billion in premiums during 2024, according to Risk Management Advisors, with projections indicating continued expansion to $135 billion by 2030 at a compound annual growth rate of 7.57 percent. North America maintains the largest regional share, while Asia Pacific demonstrates the fastest growth trajectory as mid-market companies increasingly adopt captive structures for risk management.
Captive formations continued outpacing closures for the fourth consecutive year in 2024, with nearly three new formations for every closure. Vermont licensed 26 new captives by mid-2024, bringing its total to 690 captive entities. Utah added 11 captives to reach 422, while North Carolina grew to 318 licensed captives. Cell captives dominated new formations at 38 percent of the market, with group and single-parent captives each accounting for 29 percent.
Several structural factors drive sustained growth beyond cyclical market conditions. Commercial property rates remained elevated despite some quarterly relief, while casualty markets deteriorated due to social inflation and third-party litigation funding. Natural catastrophe losses and secondary perils maintained pressure on traditional capacity. Healthcare costs continued rising, pushing more employers toward captive structures for medical stop-loss coverage. The result is a persistent demand for alternative risk financing across diverse industries and coverage types.
Technology Platforms Challenge Established Providers
Matcha enters a competitive landscape dominated by established enterprise software vendors and specialized service providers. Current solutions for risk management and captive administration remain heavily manual, fragmented across multiple platforms, and built on outdated technology architectures. These incumbents offer mature platforms with extensive client bases and deep industry relationships.
Yet traditional providers built their systems around human-intensive workflows and consulting engagements. Manual data entry, spreadsheet reconciliation, and sequential review processes characterize typical captive operations. Matcha's automation-first architecture eliminates many touchpoints that drive overhead in conventional arrangements. The platform integrates directly with existing systems through APIs rather than requiring proprietary data formats or closed ecosystems.
The strategic differentiation extends beyond workflow automation to economic alignment. Traditional captive managers and third-party administrators typically operate on service fees linked to administrative scope and activity levels. Matcha's model focuses on software delivery and AI infrastructure that reduces the cost and effort of running day-to-day operations, helping programs run more efficiently and capture more value from improved decision-making. Rather than displacing captive managers and TPAs, Matcha is built to work alongside them, strengthening their ability to deliver high-quality service at scale.
Critics question whether agentic AI possesses sufficient maturity for regulated insurance operations. Gartner predicts more than 40 percent of agentic AI projects will be canceled by the end of 2027 due to escalating costs, unclear business value, or inadequate risk controls. "Most agentic AI projects right now are early-stage experiments or proof of concepts that are mostly driven by hype and are often misapplied," states Anushree Verma, Senior Director Analyst at Gartner. She warns that many use cases positioned as agentic today fail to require agentic implementations, leading organizations to overinvest in complex technology for problems that simpler solutions could address.
The criticism highlights legitimate implementation challenges around cost management, technical complexity, and governance frameworks. Legacy system integration often disrupts existing workflows and demands costly modifications. Many vendors engage in what Gartner terms "agent washing," rebranding existing products without substantial agentic capabilities. Of thousands of vendors claiming agentic AI functionality, Gartner estimates only about 130 deliver genuine autonomous decision-making systems.
Compliance Architecture Meets Autonomous Systems
Kasian's response centers on architectural choices that prioritize regulatory alignment from initial design. Every AI decision within Matcha's platform generates traceable records that document inputs, model outputs, and final determinations. Human oversight remains embedded at material decision points, allowing underwriters and risk managers to review and override automated recommendations when circumstances warrant intervention.
The compliance-first methodology combines structured risk schemas with large language model capabilities, creating systems that leverage AI's processing power while maintaining the auditability that regulators and clients demand. Matcha's agents handle data ingestion, document classification, and preliminary analysis, then surface relevant information to human decision-makers rather than operating entirely autonomously. The approach mirrors successful patterns in other regulated industries where AI augments rather than replaces expert judgment on complex cases.
Data security and auditability rank among Kasian's primary technical concerns. Captive insurance involves sensitive financial information and strategic risk decisions that organizations cannot expose to unauthorized access or unexplained algorithmic choices. Matcha implements regulatory-grade controls throughout its infrastructure, including encryption, access management, and comprehensive activity logging. The platform maintains detailed records suitable for audits, regulatory examinations, and internal reviews.
Integration capabilities further distinguish Matcha's technical architecture. The platform connects with existing policy administration systems, claims platforms, accounting software, and data sources through standard APIs. Organizations can adopt Matcha's workflows without abandoning previous technology investments or migrating entire datasets to proprietary formats. The interoperability reduces switching costs and allows gradual implementation across different operational areas.
Mid-Market Access to Sophisticated Risk Management
The economic implications extend beyond operational efficiency to strategic optionality for businesses previously excluded from captive structures. Traditional captive formation requires substantial upfront capital, legal fees, actuarial analysis, and ongoing administrative overhead. Smaller organizations often cannot justify these expenses despite facing insurance cost pressures similar to larger enterprises.
Matcha's automation reduces fixed costs and minimum scale requirements. Associations can sponsor captives for member groups, distributing costs across participating organizations. Vertical software platforms can integrate risk management features directly into existing products, monetizing captive capabilities without building insurance expertise internally. Mid-market companies access the same risk retention economics that Fortune 500 firms have long employed, leveling competitive dynamics in industries where insurance represents significant operating expenses.
The democratization thesis rests on whether AI-driven platforms can reliably handle the judgment-intensive aspects of captive management. Underwriting decisions, claim adjudication, and regulatory compliance all require contextual understanding and discretionary authority that pure automation struggles to replicate. Kasian acknowledges these limitations while contending that technology can address perhaps 70 to 80 percent of routine tasks, freeing human expertise to focus on exceptional cases and strategic decisions.
Early evidence from pilot implementations suggests the division of labor proves workable in practice. Matcha's AI agents process standard submissions, generate preliminary risk assessments, and populate forms with transaction data. Human captive managers review outputs, make final determinations on complex cases, and maintain relationships with regulators and service providers. The hybrid model preserves necessary oversight while eliminating bottlenecks that previously constrained captive scalability.
Geographic Expansion and International Markets
Matcha currently serves United States mid-market companies and associations, concentrating efforts on domestic domiciles where regulatory frameworks and service provider networks are most mature. The company plans expansion into broader North American markets and selected international jurisdictions within the next twelve months. Geographic strategy balances opportunity assessment against regulatory complexity across different captive domiciles.
Vermont, Utah, Bermuda, and the Cayman Islands represent established domiciles with deep infrastructure and favorable regulatory regimes. France emerged as a significant new domicile following captive-specific regulations introduced in 2023, demonstrating European market fertility for alternative risk structures. Asia Pacific shows the highest projected growth at 9.2 percent CAGR through 2033, driven by economic development, expanding corporate sectors, and increasing risk management sophistication.
Each jurisdiction presents unique regulatory requirements, tax considerations, and operational standards that Matcha's platform must accommodate. The company emphasizes data security and regulatory compliance as core technical capabilities rather than jurisdiction-specific features. Building systems that adapt to diverse regulatory frameworks positions Matcha for international expansion without requiring complete architectural overhauls for each new market.
Kasian's international background and experience deploying ML systems to multi-country user bases inform the expansion approach. The infrastructure he developed at Quarks Tech and Goodsend reached users beyond his home country, providing exposure to cross-border technical and operational challenges. Matcha's API-first architecture facilitates integration with regional service providers and local compliance systems regardless of geographic location.
Industry Transformation Through Infrastructure Innovation
The longer-term vision extends beyond operational efficiency improvements to structural changes in how industries approach risk management. Kasian envisions a future where owning risk becomes the default rather than the exception, with thousands of captives operating globally on Matcha's infrastructure. The transformation would shift hundreds of billions of dollars in premiums from traditional insurance carriers to self-insured arrangements, fundamentally altering commercial insurance economics.
Critics might view such projections as aspirational, given the current market structure and regulatory constraints. Traditional insurers provide valuable functions, including risk pooling, capital provision, and specialized expertise that captives cannot fully replicate. Most organizations lack the risk appetite or financial capacity to absorb major losses, regardless of operational cost savings from AI-driven administration. The role of traditional carriers in providing reinsurance capacity and handling catastrophic exposures seems likely to persist.
Yet incremental shifts in risk retention patterns can produce meaningful industry impact over time. Even modest increases in captive penetration affect pricing dynamics, capacity allocation, and carrier profitability in traditional markets. Technologies that reduce barriers to captive formation accelerate these trends by expanding the addressable market to include organizations previously unable to access alternative risk structures.
The parallel to broader software industry disruption patterns seems relevant. Cloud computing initially addressed edge cases and small-scale implementations before gradually displacing on-premise infrastructure across enterprises. Mobile banking started with basic transactions before evolving into comprehensive financial management platforms. Matcha's trajectory may follow a similar progression, beginning with straightforward captive operations before tackling increasingly complex risk management scenarios as technology capabilities mature.
Kasian's measured perspective acknowledges both opportunity and constraint. "Our role is to make the ecosystem work together," he explains, positioning Matcha as infrastructure that connects existing service providers rather than attempting to replace the entire captive management value chain. Licensed captive managers, actuarial firms, third-party administrators, and regulatory authorities all maintain essential functions within the ecosystem. Technology automates specific workflows and eliminates friction points while preserving necessary human expertise and oversight.
The approach reflects a pragmatic understanding of how transformation unfolds in regulated industries. Revolutionary change faces resistance from stakeholders with invested capital and established relationships. Evolutionary improvements that augment rather than displace existing capabilities gain broader acceptance while still driving meaningful efficiency gains. Matcha's emphasis on interoperability and ecosystem integration positions the platform as an enhancement rather than a threat to current industry participants.
"Businesses and industries understand their own risks better than outsiders do," Kasian concludes, returning to the core insight driving Matcha's mission. The observation captures both the economic logic of captive insurance and the technological opportunity in building infrastructure that makes sophisticated risk management accessible beyond large enterprises. Whether agentic AI can deliver on that promise at scale remains an open question, yet early momentum suggests the answer may determine how alternative risk markets evolve throughout the coming decade.
ⓒ 2026 TECHTIMES.com All rights reserved. Do not reproduce without permission.
