Ola Kolade, CEO and co-founder of Underflow Inc., has raised $3 million in seed funding to build an autonomous underwriting assistant that dismantles the manual work commercial insurance professionals have endured for decades. His startup challenges an industry that has spent billions on digitization without actually removing the repetitive tasks that consume underwriters' time.
"The industry kept trying to automate the container instead of automating the work," Kolade said. "You can digitize a filing cabinet. You can build a prettier inbox. But if the underwriter still has to open every attachment, read every page, type every data point into a system, and draft every follow-up email by hand, you have not actually automated anything. You have just given them a nicer desk to do the same manual work."
The scale of the inefficiency
The numbers back his critique. Capgemini's 2024 World Property and Casualty Insurance Report found that 41 to 43 percent of commercial underwriters' time goes to administrative activities like data entry and record keeping. Only about one-third is spent on core underwriting functions such as risk assessment and premium calculation. Accenture's research paints an even bleaker picture, estimating that the average commercial underwriter spends approximately 70 percent of their time on non-underwriting activities, representing a cumulative $85 billion to $160 billion cost over five years across the industry.
Kolade's diagnosis is that prior technology solutions addressed the wrong layer. Optical character recognition can extract text from a PDF, but it cannot compare loss runs against carrier requirements or flag contradictions between forms. "Optical character recognition can read the text on a page," Kolade said. "It cannot tell you that the loss run is two years short of what the carrier requires, or that the construction type listed on one form contradicts the information on another. That is the gap. Not reading. Comprehension."
A different approach to automation
Aurora, the product Kolade built through Underflow, connects directly to a firm's existing email environment. When a submission arrives, the system reads the email and every attachment, extracts the relevant data, and structures it into a unified record. It then performs a gap analysis, identifying exactly what is missing and why, before drafting and sending follow-up emails to the broker. If no response arrives, it follows up again. Once the file is complete, Aurora packages it and routes it to the assigned underwriter.
The assistant goes beyond intake. It writes preliminary risk assessments, flags inconsistencies across documents, identifies potential coverage opportunities, and enriches submissions with public data from property records, business registries, and corporate profiles. Kolade said the goal is not a faster version of the existing process but a better outcome: "The underwriter should receive a file that is not just complete but contextualized, with analysis already done and external data already integrated."
The retirement cliff sharpens the need
Roughly one quarter of the commercial insurance workforce is aged 55 or older, and the institutional knowledge they carry-carrier appetite familiarity, risk pattern recognition, decades of relationship context-departs with them. Kolade frames this as an unsolvable hiring problem. "The industry is losing 20 and 30 years of experience every time someone retires," he said. "You cannot hire your way out of that. The math does not work. What you can do is capture the repeatable parts of that expertise in a system that scales, so your remaining professionals can focus on the judgment calls that actually require human experience."
Skepticism and the email advantage
Insurance adoption cycles are long, and trust in new systems is hard-won. Kolade said Aurora was built to work inside email because that is where submissions happen today, avoiding the friction of a new portal or login. "We do not ask anyone to change how they send submissions. We do not require a new portal or a new login. The broker sends the email the way they always have. The difference is what happens when it arrives."
The distinction between document management and document comprehension has resonated with practitioners who have seen previous tech waves fail to deliver. Kolade wrote in Rough Notes that "speed has always determined who gets the business. Now it will determine who stays in business."
Why this matters for insurance professionals
Underwriters who spend the majority of their day on manual data entry and follow-up tasks lose time that could be spent assessing risk and writing profitable business. As autonomous AI for Insurance tools like Aurora begin to shoulder these repetitive workloads, professionals who adapt early can shift their focus to complex judgment and client relationships-areas where institutional knowledge adds the most value. The pressure to do so will only intensify as experienced colleagues retire, taking decades of unwritten expertise with them.
Your membership also unlocks:
