Insurtech Shepherd, a technology platform focused on commercial industries, has closed a $42 mn Series B round led by Intact Private Capital, with participation from Spark Capital, Costanoa Ventures, and other investors.
The fresh capital will fund team growth as demand rises for faster, more accurate underwriting in construction and other sectors built around physical assets. That demand is not slowing.
Founded in 2021, Shepherd builds an AI-driven insurance platform for builders and operators. The company says it has grown nearly 7x over the past 24 months and has now raised $67 mn in total.
Shepherd currently insures more than $400 bn in project value across over 1,500 policies for 600+ customers.
Those customers include major AI labs, chip manufacturers, hyperscalers, general contractors, specialty builders, and energy developers building the physical backbone of the AI economy.
In the early 2020s, commercial insurance moved into one of its hardest markets in decades, with premiums rising sharply through broad rate increases rather than more selective repricing.
At the same time, US investment in infrastructure and data centres surged into the hundreds of billions over the last five years, with trillions more expected. Those projects need to get built. They also need cover.
Legacy insurance systems were not built for that pace. According to Beinsure analysts, this is where newer underwriting platforms are finding room, especially in construction and infrastructure lines where speed, project visibility, and risk differentiation now matter more than the old paperwork-heavy model ever allowed.
Justin Levine, Shepherd’s chief executive and co-founder, said the AI race has moved from the cloud to the construction site. He said the insurance market meant to keep those projects moving still operates much as it did decades ago.
Every GPU cluster needs a building, every building needs to be built, every data centre needs dedicated power.
Levine said Shepherd was built on the view that commercial underwriting needed a more basic reset, not only faster execution but smarter execution too. That, in his words, is why some of the biggest names in AI infrastructure are choosing the company.
Shepherd says it was built to fix long wait times and inconsistent underwriting outcomes. Its AI platform runs the same workflows in seconds, cutting underwriting feedback times from weeks to hours. That is a meaningful change in any market. In this one, it hits harder.
The platform also pulls in real-time data from partners such as Procore and Autodesk, allowing Shepherd to assess risk faster than traditional carriers still relying on static applications.
The company says it can see how a project is actually being delivered, including incident tracking, quality inspection rates, document management, and jobsite conditions updated on a continuing basis.
With those partnerships, data flows, and its AI platform, Shepherd says it can offer differentiated pricing through its Shepherd Savings programme.
The model rewards builders using construction technology and jobsite data to show stronger safety performance and better operating discipline.
The company says clients choose it for faster, more precise, and less friction-heavy underwriting in a competitive market.
Its book includes physical infrastructure tied to leading AI and technology companies building data centres, semiconductor facilities, and energy assets. Big risks, expensive timelines, no room for sluggish underwriting.
Shepherd is also pushing toward what it describes as the most ambitious technical goal in commercial insurance, fully autonomous underwriting.
Levine said the future of commercial insurance will be behaviour-based. In his view, contractors investing in technology to build more safely and intelligently should not be priced the same way as everyone else.
The company sees a shift away from manual processes toward an AI-led model where the full workflow is handled by artificial intelligence, leaving underwriters to move out of data-entry work and into oversight roles.
Mo Mahallawy, Shepherd’s chief technology officer and co-founder, said the company mapped the same autonomy path seen in self-driving vehicles and applied it to underwriting. He said an underwriter today manually handles about 20 accounts a month and loses too much time on submissions outside appetite.
Under Shepherd’s model an underwriter would oversee 200 accounts, with AI handling intake, data enrichment, risk analysis, and pricing, while the underwriter sets strategy, handles exceptions, and focuses on the accounts worth real attention.
Shepherd says it is on a two-year path toward supervised autonomy, where underwriters act as portfolio orchestrators rather than processors. If that works at scale, the company believes capacity rises 10x.
