Kalepa's $75M Series B and the case for specialized AI in insurance
New York-based Kalepa has raised a $75 million Series B at a $470 million valuation. Updata Partners led the round, with existing backers IA Ventures and Inspired Capital participating. Kalepa announced the deal on November 5, bringing its total funding to $91 million.
The size of the jump is what stands out. Kalepa's last raise was a $14 million Series A in 2021. For five years the company barely touched outside capital, and it has been cash-flow positive for more than two of them. In a market where AI startups often raise every nine months, that restraint is a big part of why investors were willing to pay up now.
The problem Kalepa is solving
Commercial insurance underwriting is still remarkably manual. When a broker wants to place coverage for a restaurant chain, a construction firm or a warehouse, the insurer receives a "submission". That is usually an email with PDFs, spreadsheets, loss histories and application forms in no standard format.
An underwriter then has to read all of it, check whether the insurer already has the account ("clearance"), decide whether it fits appetite, research the business, price the risk and send a quote. Much of that time goes into data gathering rather than judgment. Brokers, meanwhile, send the same submission to many carriers and reward whoever replies first with a reasonable price.
Speed and accuracy pull in opposite directions. Insurers that quote fast without good analysis tend to write bad risks. Insurers that analyze carefully lose deals to faster rivals. Kalepa's whole pitch is that AI can break that trade-off.
What the platform does
Kalepa describes itself as an insurance-specific AI "harness" that coordinates AI agents across the underwriting workflow. Its platform covers seven stages: submission ingestion, clearance, triage, risk analysis, rating, quote and bind, and portfolio management.
Two design choices matter here:
Adjustable autonomy. The system takes on more or less of a task depending on how risky it is. Extracting data from a loss run can be automated end to end. Pricing a complex account keeps a human in the loop for review and approval.
Auditability. Results are verified and traceable. That is not optional in a regulated industry, where insurers must explain to regulators, reinsurers and courts why a risk was priced the way it was.
CEO and co-founder Paul Monasterio frames this as "Professional Grade AI." In his words, "Insurance runs on the judgment of experienced professionals, and our Professional Grade AI is held to that same standard." He describes the goal as letting insurers ask, task by task, "who does it best, and at what cost."
The numbers behind the round
Kalepa shared unusually specific results. Over the past year, it says revenue and customer count both tripled, and work volume on the platform grew fourfold. Customers report:
Metric | Reported result |
Premium written per underwriter | +30% to +60% |
Time to quote | 58% faster |
Quote-to-bind rate | 2x |
Combined ratio | 960 basis points better |
The last line is the one insurance executives will notice. The combined ratio is losses plus expenses as a share of premium. Below 100% means an insurer makes money on underwriting itself. A 9.6-point improvement can turn a marginal book into a very profitable one. These figures are company-reported and will vary by customer and line, but they show the kind of outcome Kalepa is selling: better risk selection, not just faster paperwork.
The customer list is serious too. Kalepa names W.R. Berkley, Bowhead Specialty and AmRisc, and shows logos including Munich Re, Church Mutual, Canopius and Secura. Selling to carriers is notoriously slow, with long pilots and heavy security reviews. Landing these names is a moat in itself.
Why vertical AI is getting funded
Kalepa fits one of the clearest patterns in insurtech funding this year: investors are backing vertical AI companies that own a full workflow in a regulated industry, rather than general-purpose tools. Horizontal AI assistants can summarize a PDF. They cannot clear a submission against a carrier's book, apply its underwriting guidelines and leave an audit trail that satisfies a regulator.
The team reflects that bet. CTO Garrett Schlesinger spent 12 years at Affirm, where he was the 12th employee and helped build its real-time underwriting system. Consumer lending and commercial insurance are different worlds, but the core problem of making fast, explainable risk decisions at scale is the same.
Growth equity firm Updata Partners leading the round also says something. Updata typically backs software companies with real revenue and efficient growth, not pre-revenue moonshots. A $470 million valuation on a capital-light, cash-flow-positive business suggests investors are paying for durability rather than hype.
The risks
Kalepa is not alone. Startups such as Sixfold are going after the same underwriting workflow. Incumbent policy-administration vendors and large carriers' internal AI teams are building similar tools. Big insurers in particular may prefer to build in-house once they see what is possible.
There is also model risk. If AI-assisted underwriting systematically misprices a class of business, the losses may not show up for years, especially in long-tail lines such as liability. Insurers will want years of loss data before they fully trust any automated risk selection.
Finally, scaling a services-heavy enterprise sale is hard. Kalepa plans to double headcount in 2027. Doing that while keeping the efficiency that made it attractive will be a test of execution.
What to watch
Kalepa says the money will fund new AI products, expansion into new markets and that hiring push. The signals worth tracking:
Whether it expands beyond US commercial lines into Europe, the London market or reinsurance, where partners like Munich Re and Canopius already sit.
New customer wins among top-20 US commercial carriers.
Independent or carrier-published evidence of loss-ratio improvement over a full underwriting cycle.
Whether it stays cash-flow positive while doubling the team.
If those hold, Kalepa will be a strong argument that the most valuable AI companies of this cycle may be the quiet, specialized ones.