Insurance tech startup Shepherd raises $42M to underwrite the physical layer of AI – News

Artificial intelligence-native insurance provider Shepherd has just raised $42 million in a new funding round to underwrite the massive infrastructure boom that’s fueling the industry’s growth.

The Series B round was led by Intact Private Capital and saw participation from Spark Capital, Costanoa Ventures and others, bringing the company’s total amount raised to $67 million.

Cloud hyperscalers, chipmakers and frontier AI labs have collectively poured hundreds of billions of dollars into building out the infrastructure needed to support AI’s growth, such as data centers, power plants and chip fabrication plants. But a bottleneck has emerged that’s holding back many of those projects: They all need to be insured before they can get up and running.

Insurance is not an exciting industry, but it is essential to the AI boom. Contractors are legally obliged to have insurance policies in place before they can break ground on a new multibillion-dollar data center, but that has proven to be a difficult issue.

The problem is that the insurance industry is notoriously slow, moving at a snail’s pace as carriers carefully analyze each project using disconnected tools, in order to come up with a quote. For the contractor, who needs to shop around and obtain multiple quotes, it usually means weeks of making phone calls, sending emails and chasing people up.

Shepherd co-founder and Chief Executive Justin Levine (pictured, left) argues the entire process is completely outdated in the AI era, where everything is supposed to move so much faster. “The AI race has moved from the cloud to the construction site,” he said. “Every GPU cluster needs a building, and the insurance market that is supposed to keep those projects moving has been operating the same way for decades.”

As is the case with almost everything these days, Shepherd’s solution is to use AI to speed things up. It says its AI-native platform collapses the multiweek underwriting process into just a few hours by directly integrating with the “source of truth” on construction sites – namely, software platforms such as Autodesk, OpenSpace, DroneDeploy, Procore and Samsara.

Those third-party platforms are widely used by contractors up and down the U.S., and typically contain all of the plans and construction details for complex infrastructure projects. By integrating with them, Shepherd can unleash its AI agents to automate the analysis process, enabling it to come up with a quote in just a few hours.

The company also differentiates itself through behavior-based pricing through its “Shepherd Savings” program. Similar to how telematics can reduce the cost of car insurance premiums if the owner drives safely and responsibly, Shepherd taps onsite data to reward builders that take safety seriously. So if a contractor uses technology to track incidents and perform quality inspections, for example, it can reduce its risk profile and benefit from lower costs. Builders that invest in safety technology can therefore realize significant savings.

Shepherd’s platform has a lot of appeal. The startup says it has experienced “hypergrowth” over the last two years, growing its revenue by more than seven times, off an undisclosed base. To date, it has insured more than $400 billion in project value through more than 1,500 policies. It has more than 600 customers, including some major chipmakers, hyperscale data center operators and renewable energy firms.

The company doesn’t intend to rest on its laurels, though. With today’s funding, it will be able to progress further toward its goal of fully autonomous underwriting and accelerate the approval process. At present, it still relies on human underwriters, who are capable of handling about 20 accounts per month. But by using AI agents to handle intake, data enrichment and risk analysis, underwriters could oversee more than 200 accounts at once.

“We are closing in on the first fully agentic submission in commercial insurance,” said Shepherd’s other co-founder, Mo Mahallay (right). “Email in, price out. No human intervention until the last mile.”

Photo: Shepherd