When architects design a commercial building, they usually turn the design over to a specialized consulting firm to plan the heating and air conditioning, mechanical, electrical, and plumbing systems. Those consultants have for some time used computer-assisted design software, but it is still a laborious process that can take a skilled building engineer weeks to complete.
Now a Swedish startup called Endra is using AI to automate this mechanical, electrical, and plumbing (MEP) design process. Niklas Lindgren, the company’s cofounder and CEO, says its software can reduce the time needed to design a code-compliant electrical system for a 500,000-square-foot commercial building from two months to less than a day.
Endra is announcing today that has secured a $20 million seed round, one of the largest funding rounds to date for a Swedish company. The investment is being led by venture capital firm Notion Capital, with participation from Norrsken VC and angel investors. The company had previously raised a €3 million ($3.4 million) pre-seed round in May.
The company, which Lindgren founded roughly a year ago, currently employs about 10 people. The new funding will support its plans to triple the team size within six to 12 months, and to establish offices in the United States, United Kingdom, and Germany, the company said.
The idea for Endra grew out of serial entrepreneur Lindgren’s work at a company he previously cofounded, Sectragon, which made physical access control and security systems. He sold it to a private equity firm in 2022. While at Sectragon, Lindgren and cofounder Anton Juric, who served as Sectragon’s CEO, watched teams working on security system blueprints, and noticed the antiquated software firms used to map out the electrical systems for commercial real estate. After they sold Sectragon, Lindgren and Juric decided that their next company would focus on automating this work.
Lindgren’s father, an architect, also told him about the highly-manual process used by engineering design firms to work up the plans for HVAC and plumbing networks, which often represented a bottleneck in the building design process. “Nothing has really happened, from a technology standpoint, for the last 25 years,” Lindgren tells Fortune. “We’re building a purpose-built platform for the MEP design engineering industry, speeding up processes for human designers by hundreds of times.”
Juric cofounded the company with Lindgren and serves as its chief operating officer. Other cofounders include David Rydberg and Gustav Hammarlund, both of whom focus on building Endra’s core technology. Right now, Endra’s system only does electrical network design, but the company plans to add HVAC layout planning and plumbing network mapping within the next year, Lindgren says.
Endra integrates with Autodesk’s Revit, the standard computer-aided design software tool used across the MEP design industry. Engineers upload a three-dimensional architectural model into Endra, select the building code requirements and system type they want to design—such as lighting fixtures, power distribution, or fire alarms—and the platform generates the engineering model along with the necessary calculations and documentation.
Endra’s system is a hybrid one, Lindgren said, combining large language models (LLMs) with other forms of machine learning, 3D simulation methods, and deterministic algorithms. LLMs are used to understand information about what a building’s function is and to extract design parameters from the building specs. “They help us identify room types, identify objects,” Lindgren said, “and also the intent of the architect, like how the architect thinks this building should be used.”
With that information in hand, Endra’s system uses its own 3D-modeling algorithm and deterministic software to figure out the optimal way of arranging electrical wiring while also meeting the building codes for the relevant country or region.
The company launched with its first customers in August and has since accumulated a waiting list of more than 600 companies from over 90 countries. The rapid demand reflects broader pressures on the construction industry, where productivity has stagnated while labor shortages have intensified.
Lindgren acknowledged that venture capitalists often ask him about the startup’s competitive moat, and whether a company such as Autodesk could replicate Endra’s technology. He pointed to the company’s proprietary geometry engine and data model as differentiators, noting that replicating them would require substantial changes to existing systems.
