Kalepa announced a $75 million Series B round at a $470 million valuation on November 5, led by Updata Partners with IA Ventures and Inspired Capital participating. The insurance-AI company plans to expand its products, enter new markets and double headcount in 2027. It says total capital raised has reached $91 million.
Behind the financing is a less glamorous task than generating pictures or chat: turning documents and loss histories into underwriting decisions that can be explained. Kalepa aims to connect work from submission intake through pricing and portfolio management. According to the company, agent autonomy varies with the task and its risk, rather than providing blanket permission to replace professional judgment.
From an application to a decision-ready file
Kalepa's product site describes a sequence beginning with document classification and extracting information from forms, property schedules and loss records. It then checks completeness, matches applications to the insurer's criteria and prioritises opportunities. The aim is to reduce assembly work so underwriters can focus on the risk and the terms on which their company will accept it.
In an interview with Finextra, co-founder and chief executive Paul Monasterio said the platform can reconcile applications with loss histories, identify missing or inconsistent information and apply insurers' underwriting and pricing rules. He said it can prioritise submissions and prepare quotes within the authority granted by the insurer. Exceptions go to a professional with supporting evidence.
Reading a document and making a decision are different tasks. A tool may extract an address and insured amount correctly without knowing whether the evidence is sufficient for pricing. When details are missing, the issue is whether the process stops and refers the case to someone authorised to decide. Those boundaries must remain clear when several agents operate in sequence.
Kalepa says its system combines models with deterministic software, meaning explicitly defined rules, and adapts them to each insurer. It says outputs are checked and evidence and reasoning retained. These are the supplier's descriptions of its platform, not proof that every automated decision is accurate or appropriate for every insurance category.
Growth figures come from the company
Kalepa said revenue and customer numbers tripled over the past year and that it has remained cash-flow positive for more than two years. Platform workload grew fourfold, it added. These are separate metrics: revenue, customer counts and activity can change at different rates within the same business.
The company names W.R. Berkley, Bowhead Specialty and AmRisc among customers. It reports that users achieved 30% to 60% more premium per underwriter and a 58% reduction in quotation time. Those are company-reported results, not guaranteed outcomes for a new customer. More premium per employee alone does not establish a more profitable insurance portfolio.
Underwriting also concerns risks that emerge after a policy is sold. Processing speed is therefore only part of the picture. Teams evaluating a system need to examine conflicting information, decision reconstruction and changes to company rules. A short test with straightforward documents may not represent the exceptional submissions that arrive later.
Monasterio described a division of labour in which AI performs preparation while people make consequential decisions. He said that division can evolve as systems demonstrate reliability on specific tasks. The financing announcement therefore does not mean removing underwriters from the process. Kalepa describes matching autonomy to consequences and leaving exceptions to professional judgment.
Expanding across the insurance workflow
Kalepa was founded in 2018. Its previous $14 million round took place in 2021, led by Inspired Capital with IA Ventures participating. The new round adds $75 million, while the $470 million valuation is a separate number: an assessment of company value in the transaction, not cash received, annual revenue or the value of policies managed.
Expansion plans include insurance product development, reinsurance and distribution alongside existing operations, underwriting, pricing and portfolio management. Reuters reported Monasterio saying most hiring would be in engineering, AI and commercial roles. Doubling the workforce in 2027 is a plan, not a statement that those people have already joined or every product is available.
Expansion also raises the importance of continuity between systems. A detail entered in an application can inform risk assessment, pricing and later analysis of the whole portfolio. If wrong, automated transfer can propagate the mistake. Returning to the original document and identifying where a decision was made matter alongside savings in typing and switching screens.
Kalepa's site also describes integrating pricing models from Excel, internal systems and external rating tools. Connecting existing rules can be central to implementation. A new system does not begin with a blank slate: it enters a network of authorities, documents, procedures and historical information, all affecting how a quote becomes a policy.
What will test the promise after funding
The potential value is not merely a faster answer but a working file showing what is known, what is missing and why an action was proposed. When human approval is needed, the professional must see the supporting material. Otherwise preparation time saved can become additional review and reconstruction time precisely where mistakes are costly.
For insurance, the announcement adds investment in AI designed to operate inside defined workflows. For Kalepa, the next task is expanding coverage while preserving the boundary between delegable work and decisions requiring experience and authority. Daily use will test that distinction through information quality, exception handling and the ability to explain a decision after a quote has been sent.