On the same day Build announced its $8.5m seed round, a hospital trust somewhere in England was almost certainly submitting a business case for capital funding it would not see approved for months, if at all. That juxtaposition is not accidental. Build exists because infrastructure development, the unglamorous work of site selection, due diligence and planning coordination, has become one of the most expensive bottlenecks in the modern economy. Its investors, who include senior figures from OpenAI, Meta and Blackstone, are betting that AI can compress timelines that have stayed stubbornly slow even as computing power and capital have both grown cheaper.

The company's pitch is straightforward. Consulting and engineering firms lack the margins to invest properly in digital transformation, so projects stay trapped in fragmented, manual workflows that can add weeks or months to delivery. Build's answer is to insert an AI layer that automates the due diligence and site assessment work that currently eats up planning time. It is a bet on process, not physics.

Britain's NHS estate is arguably the starkest illustration of exactly the problem Build claims to solve, and also the clearest warning about its limits. The New Hospital Programme has spent years absorbing delay after delay, its delivery dates repeatedly pushed back as costs rose and political attention shifted. This summer's heatwave did not create the vulnerabilities in NHS infrastructure so much as expose ones that capital planners have known about for years: ageing estate, RAAC-affected buildings, cooling systems never designed for the temperatures they now face. None of this is a mystery to NHS leadership. It is a financing and sequencing problem that outpaces the organisation's ability to act on what it already knows.

That is where the comparison with Build becomes instructive rather than flattering. The startup's thesis assumes that workflow friction, the fragmented handoffs between planners, engineers and investors, is the primary constraint on infrastructure speed. Strip that away with better tooling and projects move faster. For private developers building data centres or power infrastructure, where capital allocation decisions sit largely within one organisation, that logic holds up reasonably well. The NHS's capital problem is structurally different. Its constraint is not primarily technical coordination but a chain of Treasury sign-off, political risk aversion and a procurement culture built to withstand scrutiny rather than move quickly. An AI system that speeds up due diligence does nothing to change a Spending Review cycle or a minister's appetite for headline capital announcements.

This matters for how NHS leaders and DHSC officials should read the current wave of AI-infrastructure enthusiasm coming out of the private sector. There is a real opportunity here, particularly around the more mechanical parts of estate planning: condition surveys, backlog maintenance triage, procurement documentation that currently consumes disproportionate staff time. Sir Jim Mackey's push for tighter contract accountability across trusts is precisely the kind of environment where faster, more consistent due diligence tooling could earn its keep. But importing Silicon Valley's framing wholesale, that infrastructure delay is fundamentally a software problem, risks flattering a diagnosis that lets the harder political questions go unanswered.

Andy Burnham's devolutionary instincts add another layer worth watching. If capital decisions move further from Whitehall and closer to regional health structures, the appetite for tools that promise faster site assessment and planning coordination may grow, particularly among newly consolidated ICBs trying to make thirty years of estate backlog legible in a shorter time frame. Whether that appetite translates into procurement is a different question, one that depends more on NHS England's residual capital rules, post-abolition, than on anything Build or its rivals can build.

The company's investors are wagering that the world's infrastructure gap is, at root, a coordination failure that AI can dissolve. The NHS estate suggests a more sobering possibility: sometimes the coordination failure is the point, a deliberate, if costly, feature of how public money gets spent under scrutiny. Faster due diligence will not fix that. Only a willingness to spend differently will.