Hyperexponential (hx), a technology company that develops underwriting and pricing software for commercial property and casualty (P&C) insurers, has introduced Hyperoperator, an AI-powered underwriting agent designed to progress broker submissions from initial receipt to a decision-ready file while operating within an insurer’s existing pricing models, underwriting appetite and delegated authority framework.
The company said the launch positions hx as the first provider of an agentic underwriting workbench for commercial P&C insurance capable of processing a broker email through to a fully priced and quoted risk.
According to hyperexponential, conventional underwriting workbenches primarily help underwriters organise and manage their workload. By comparison, the company’s agentic underwriting workbench is intended to complete routine preparation and administrative tasks before referring cases to underwriters for decisions that require professional judgement.
Hyperexponential said Hyperoperator integrates with the hx platform, email, Microsoft Teams, Slack and mobile devices. Underwriters can begin a workflow by forwarding a broker submission, after which the system extracts and structures relevant information, determines the next permitted step, applies the insurer’s underwriting rules and prepares a file for review. Depending on the insurer’s configured controls, submissions can either be routed to an underwriter or continue automatically through the underwriting process.
The company said commercial P&C insurers are under increasing pressure to provide faster responses to brokers while maintaining underwriting discipline and profitability in softer market conditions. Although many insurers have trialled AI technologies, hyperexponential believes relatively few have embedded them into live underwriting operations.
Hyperexponential noted the main obstacle has not been the capability of AI itself, but the fragmented nature of underwriting knowledge, including pricing models, appetite guidelines, authority limits, portfolio insights and institutional knowledge spread across multiple systems and documents. As a result, the company stated many AI initiatives have been confined to submission intake or triage rather than supporting the underwriting decision itself. Hyperoperator, according to hyperexponential, has been developed to work within each insurer’s underwriting strategy, governance framework and decision-making processes rather than alongside them.
Speaking during a keynote presentation at hx Live London, Amrit Santhirasenan, CEO and Co-Founder of hyperexponential, demonstrated how Hyperoperator coordinates specialist AI agents across the hx platform to prepare underwriting decisions using an insurer’s existing data, models and controls. Hyperexponential said a live demonstration showed the platform processing a cyber insurance submission from broker email to a triaged and priced risk in less than three minutes without manual data entry.
“For years, underwriting software has helped teams manage the work around the decision,” added Santhirasenan. “The opportunity now is to bring agents into the decision workflow itself, with the pricing logic, portfolio context, and authority controls that serious underwriting requires. hyperoperator gives underwriters leverage without taking judgment out of their hands.”
According to Hyperexponential, the platform is used by leading global insurers to support underwriting decisions representing more than $75 billion in annual commercial P&C premium, compared with $45 billion a year earlier. The company said this reflects eight years of experience supporting production underwriting environments.
Hyperoperator builds on the company’s existing AI capabilities for submission ingestion, model development, model testing and risk assessment. Those technologies, which are already deployed across the hx customer base, have now been brought together to support the wider underwriting process from submission through to decision.
This approach is intended to ensure each submission receives an appropriate level of review, while helping brokers receive faster responses and enabling underwriters to focus more of their time on technical judgement rather than administrative work.
The company said insurers can deploy Hyperoperator in several ways, including fully automated processing where confidence levels and authority limits allow, workflows requiring underwriter approval, on-demand underwriting analysis during live cases, and continuous monitoring of portfolio-level indicators such as exposure concentrations, pricing adequacy and unusual risk patterns.
“Most carriers believe they are further along the AI maturity curve than they actually are,” said Karlyn Carnahan, Head of Insurance North America at Celent. “The next phase of AI in insurance will not be defined by isolated pilots. It will be defined by how safely and effectively carriers bring AI into the workflows where underwriting decisions are prepared, reviewed, and improved over time.”
The hx platform combines workflow management, underwriting rules, rating engines, approval processes and portfolio intelligence within a governed operating environment. As Hyperoperator operates through this infrastructure, the company said its AI agents remain subject to insurer-defined governance, approval pathways and underwriting controls throughout the decision-making process.
“Most AI in underwriting stops at the first 20% of the workflow: reading submissions, routing tasks, and helping teams move faster,” noted Marcus Ryu, General Partner at Battery Ventures and Co-Founder of Guidewire. “The hard part is the other 80%: connecting AI to the pricing logic, workflow controls, portfolio context, and trust infrastructure that underwriting decisions depend on. hyperoperator is compelling because hx has spent years building that foundation in production.”
