The expansion reflects a broader shift in AI investment from general-purpose models to industry-specific applications.
AI is increasingly moving beyond software development and office productivity into one of the construction industry's most labor-intensive disciplines: mechanical, electrical, and plumbing (MEP) engineering.
Stockholm-based AI startup Endra announced on July 31 that it has expanded into the United States and the United Kingdom, establishing offices in New York, San Francisco, and London. The move follows the company's $50 million Series A financing led by Andreessen Horowitz last month.
The expansion comes as global construction activity accelerates, driven by investment in AI data centers, power infrastructure, healthcare facilities, semiconductor manufacturing, and low-carbon buildings. Industry participants say engineering capacity is becoming one of the biggest constraints on project delivery.
A Critical but Often Overlooked Part of Construction
MEP engineering forms the technical backbone of modern buildings, covering electrical systems, HVAC, plumbing, fire protection, and other essential infrastructure. Although these engineering services typically account for only 1–5% of total construction costs, they influence virtually every stage of a project's execution.
Delays in MEP design often affect procurement, construction schedules, and project completion, making engineering capacity a critical constraint despite representing a relatively small share of overall project budgets.
Industry estimates place the global MEP services market at more than $150 billion annually, with demand continuing to grow as buildings become more technologically complex.
Automation Targets Engineering Bottlenecks
Despite advances in digital design, many engineering firms continue to rely heavily on manual workflows for system design, documentation, and multidisciplinary coordination.
Endra develops an AI platform designed to automate many of these repetitive engineering tasks while integrating with existing Building Information Modeling (BIM) software, including Autodesk Revit. Rather than replacing established engineering workflows, the platform is intended to reduce repetitive work so engineers can spend more time on technical decision-making and complex design challenges.
According to the company, its software can reduce certain engineering design processes from several weeks to less than a day. One example cited by Endra is electrical design for a 500,000-square-foot commercial building. The performance figures have not been independently verified.
Construction Becomes a New Frontier for Vertical AI
The significance of Endra's expansion extends beyond the company's geographic growth.
For much of the past two years, AI competition has focused on general-purpose models powering chatbots, search engines, coding assistants, and enterprise productivity tools. Increasingly, however, investment is shifting toward vertical AI—industry-specific systems designed to solve specialized operational problems.
Construction is emerging as one of those markets.
Engineering firms worldwide continue to face shortages of experienced MEP engineers while demand rises for increasingly complex facilities such as AI data centers and electrified infrastructure. As projects become larger and more technically demanding, improving engineering productivity has become a strategic priority rather than simply an efficiency gain.
The trend also reflects growing investor interest in AI companies addressing physical-world industries rather than digital workflows alone.
Expansion Follows Rising Infrastructure Investment
Endra said its U.S. expansion will be led from New York, while its London office will support engineering consultancies across the United Kingdom, Europe, and the Middle East.
The company plans to continue expanding its engineering and commercial teams as it prepares for Epoch, its product event scheduled for September in Las Vegas.
As governments and technology companies accelerate investment in AI infrastructure and energy systems through the end of the decade, demand for engineering automation is expected to continue rising.
The next phase of AI competition is increasingly being defined not by broader general-purpose models, but by vertical AI that removes bottlenecks in industries such as construction, healthcare, manufacturing, and finance.
