- James Stirrat-Ellis, co-founder of Build, trained as an architect and worked on the S$13 billion Changi Airport expansion. He later developed AI to speed up the due diligence process that often delays infrastructure projects worldwide.
- The New York-based startup has raised $8.5 million in seed funding, led by Index Ventures. Angel investors from OpenAI, Blackstone, and Meta AI Research also joined the round.
- Build has completed more than 100 projects across 15 countries and recently landed its first hyperscaler client, referring to the largest cloud and tech companies.
While working on the $13 billion Singapore airport terminal, which was built to serve 50 million passengers a year, James Stirrat-Ellis saw major problems such as lengthy consultant reports, scattered data, and slow workflows.
“I realised I either need to start my own firm or get the tech to build things faster,” he tells Tech Funding News. He left his master’s program at Harvard, taught himself to code, and helped lead a fast-growing AI company in 2024 before starting Build, an AI infrastructure company, with Ben McClusky.
Today, this New York-based startup raised $8.5 million in seed funding led by Index Ventures. Pebblebed, Puzzle Ventures, and Tiny.vc also joined, along with angel investors such as OpenAI CFO Sarah Friar, Blackstone CTO John Stecher, and senior leaders from Meta AI Research and Google Maps.
Focusing on services rather than software
Unlike most AI startups that sell licenses, Build works on a retainer basis. This difference is central to how the company operates.
Infrastructure projects often rely on consultants for tasks such as site sourcing, technical checks, power assessments, risk planning, and environmental analysis. These jobs are usually handled by different firms over several weeks. With hyperscalers expected to invest almost $700 billion in data centres by 2026, speed is essential.
Build cuts out these consulting costs. Clients pay a monthly retainer and get integrated assessments from over 1,600 data sources, checked in parallel and reviewed by a human before delivery.
“The industry spends money on consultants and outsourced labour. We replace that spending. We achieve software margins by automating much of our work, but our focus is on services spend, not software spend,” Stirrat-Ellis says.
The company says this approach cuts due diligence timelines by more than 95%. Tasks that once took weeks now take only hours. This model is catching on in real estate and property tech, where AI is starting to automate jobs that people used to handle.
Growing from two founders to serving major clients in under two years
Since its inception, the company has grown to 10 employees across four continents and plans to hire five more. It has completed over 100 projects in 15 countries for clients such as Tishman Speyer, supporting firms with a total of $2 trillion in assets. Build also recently signed its first hyperscaler client, which Stirrat-Ellis calls “the top customer you can get” in data centre work.
This growth shows a gap in the market. By the end of 2025, global data centre capacity under construction will hit 23 gigawatts. Demand driven by AI is outpacing what traditional development can keep up with. While big consultancies like CBRE, Arup, and JLL are investing in AI tools, Build stands out by replacing consultants instead of just making them more efficient.
“In terms of direct technology competitors, we were first to market. We compete with the legacy industry, the consultants that existed before,” Stirrat-Ellis says.
This approach caught the attention of Index Ventures, known for backing Revolut, Deliveroo, and Adyen, with partner Martin Mignot leading many investments. “The round came to us. We got a term sheet within a day,” Stirrat-Ellis tells TFN.
“James and Ben have combined deep technical capability with real operational understanding of how critical infrastructure gets built. The infrastructure challenge is global, and what the team has achieved before launch is extraordinary,” adds Mignot.
Moving toward agentic real estate
The fresh funds will go toward two main goals: hiring engineers to improve the technology and bringing in experts from firms like Blackstone, Starwood Capital, and JP Morgan to help connect AI results with client needs.
Build aims to automate the whole development process over the long term, from site selection and permits to engineering, pre-construction, and asset management. They call this approach agentic real estate. Revenue is growing, and the next step is to quickly grow the customer base.
“The industries shaping the physical world have spent decades trapped in process instead of creativity. By removing that operational burden, we can help teams move faster, make better decisions, and deliver better infrastructure. That’s the long-term opportunity, and we’re only at the beginning of it,” Stirrat-Ellis concludes.
The main question is whether Build can keep its lead over time. Its current strengths are speed, focus, and a business model that others have not copied yet. As big consultancies invest more in AI, the future of both the company and the industry will depend on keeping this advantage.
