The rapid expansion of artificial intelligence isn’t happening solely in the cloud. It’s increasingly reliant on a massive build-out of physical infrastructure – data centers, specialized chip fabrication plants, and the power grids to support them. And as investment in this “physical layer of AI” surges, a new bottleneck has emerged: insurance. A San Francisco-based insurtech startup, Shepherd, is aiming to solve that problem, recently raising $42 million in a Series B funding round to streamline insurance for these complex construction projects.
The funding round, led by Intact Private Capital with participation from Spark Capital and Costanoa Ventures, brings the company’s total funding to $67 million, according to a press release. Shepherd’s core proposition is to drastically reduce the time and complexity involved in securing insurance for large-scale AI infrastructure projects, a process that traditionally takes weeks and involves a significant amount of manual work.
The demand for insurance is driven by the sheer scale of investment in AI infrastructure. Cloud providers, chip manufacturers, and AI labs are collectively pouring hundreds of billions of dollars into new facilities. “Every GPU cluster needs a building,” explains Justin Levine, Shepherd’s co-founder and CEO. “And the insurance market that is supposed to retain those projects moving has been operating the same way for decades.” Contractors are legally required to have insurance in place before breaking ground on these projects, but obtaining quotes from traditional insurance carriers has become a major impediment to progress.
The Slow Pace of Traditional Underwriting
The traditional insurance underwriting process is notoriously slow. Carriers typically rely on manual analysis of project details, often using disconnected tools and requiring extensive back-and-forth communication with contractors. This means weeks of phone calls, emails, and follow-ups just to obtain a quote. The problem isn’t a lack of willingness to insure these projects, but rather the inefficiency of the existing system in handling their complexity and the speed at which they are being proposed.
Shepherd’s solution leverages artificial intelligence to accelerate this process. The company’s platform integrates directly with the software already used on construction sites – including Autodesk, OpenSpace, DroneDeploy, Procore, and Samsara – to access project plans and construction details. By connecting to these “sources of truth,” Shepherd’s AI agents can automate much of the data analysis traditionally performed by human underwriters.
AI-Powered Underwriting and Behavior-Based Pricing
According to Shepherd, this integration allows them to collapse the multi-week underwriting process into a matter of hours. The AI agents handle intake, data enrichment, and risk analysis, freeing up human underwriters to focus on more complex cases. Currently, human underwriters at Shepherd can handle approximately 20 accounts per month. The company aims to increase that to over 200 accounts per underwriter as the AI agents take on more of the initial workload.
Beyond speed, Shepherd differentiates itself through a program called “Shepherd Savings,” which offers behavior-based pricing. Similar to how telematics data can lower car insurance premiums for safe drivers, Shepherd rewards builders who prioritize safety on the job site. By tracking incidents and quality inspections using technology, contractors can demonstrate a lower risk profile and potentially reduce their insurance costs. This incentivizes the adoption of safety technologies and promotes safer construction practices.
Rapid Growth and Future Plans
Shepherd reports significant growth over the past two years, with revenue increasing more than sevenfold, though the company has not disclosed its base revenue figure. To date, the platform has insured over $400 billion in project value across more than 1,500 policies, serving over 600 customers including major chipmakers, hyperscale data center operators, and renewable energy firms.
Mo Mahallay, Shepherd’s other co-founder, envisions a future where insurance submissions are almost entirely automated. “We are closing in on the first fully agentic submission in commercial insurance,” he said. “Email in, price out. No human intervention until the last mile.”
With the new funding, Shepherd plans to accelerate its progress toward fully autonomous underwriting and further refine its AI-powered platform. The company is focused on expanding its integrations with construction software and enhancing its risk assessment capabilities. The ultimate goal is to provide a seamless and efficient insurance experience for the rapidly growing AI infrastructure sector.
The increasing demand for AI infrastructure, coupled with the complexities of insuring these projects, suggests a significant opportunity for companies like Shepherd. As the AI boom continues, the demand for faster, more efficient insurance solutions will only become more critical. Shepherd’s next step will be to demonstrate the full potential of its autonomous underwriting capabilities and further scale its platform to meet the growing demands of the industry.
This article provides information for general knowledge and informational purposes only, and does not constitute professional advice.
Share your thoughts on the evolving intersection of AI and insurance in the comments below.
