Unlimited Industries’ AI-Powered Construction Raises $12 Million Seed Round

When Andreessen Horowitz and CIV committed to leading the $12 million seed round for Unlimited Industries, the move would have raised eyebrows in a pre-AI world. But the pattern of AI reinventing legacy sectors from accounting receivables to healthcare admin has become so familiar that backing a construction firm now fits a “New Industrials” trend where firms take on foundational physical systems with software-like execution, combining urgency with scale broader logic.

Founder and CEO Alex Modon says the industries Unlimited is targeting “are deeply frustrated with how slow and costly their projects move today, and haven’t experienced what modern AI can unlock.” That gap is exactly what drew him to this vertical. A multidisciplinary engineer who previously worked scaling B2B SaaS and commercializing deep-tech alongside U.S. DOE labs, Modon founded Unlimited after repeatedly running into the limits of legacy construction workflows in his previous roles.

Unlimited combines an AI-powered design platform with its own integrated engineering and construction teams, offering fixed-price contracts instead of the traditional cost-plus model. The platform generates and evaluates tens of thousands of design configurations in parallel to surface layouts optimized for cost, safety and performance before ground is ever broken. The result can compress pre-construction timelines from months into weeks. The company reports a recent case where projected capital costs were reduced by more than half through an AI-driven redesign.

The startup’s leadership team combines heavy-industry credibility and product-engineering chops: Co-founders Jordan Stern and Tara Viswanathan previously founded and scaled a high-growth startup to a nine-figure exit. Unlimited’s manifesto frames its mission in bold terms: infrastructure shouldn’t just maintain; it should inspire. “When it comes time to build, robotics with embodied intelligence will bring that design into reality at massively accelerated speed and with incredible precision,” the company writes on its website.

In our conversation, Modon returned repeatedly to the industry’s inertia with doing business the same old way. “A lot of these people have not had their ChatGPT moment,” he said. “They are looking at workflows that have not changed in decades. They have no priors for what this technology can actually do.” This makes product demonstrations unusually effective. When Unlimited shows an engineering team a multivariate design search that explores tens of thousands of configurations, the reaction is immediate. “It looks like magic to them,” Modon said.

Unlimited’s early customers include young energy technology companies and large industry incumbents that both need projects delivered far faster than existing contractors can support. These buyers care about acceleration more than anything. A data center developer that waits six or nine extra months before breaking ground loses revenue, market share, and power capacity it may never recover. If you’re making a connection between Energy, Data Centers, Unlimited, and Andressen’s role nurturing an ecosystem of AI startups, addressing every niche as quickly as they can be found and matched with a team.

Unlimited currently has a handful of live projects across mining, power, and data infrastructure. The total project values range from a quarter to half a billion dollars, with Unlimited’s portion structured as fixed-price work, instead of cumbersome change orders and continual re-pricing that slows the process today. He argues that removing change orders is fundamental. “We do changes all the time,” he said. “That is the point. But we do not let the contract punish the customer for learning. Iteration should be natural, not a revenue event.”

“The engineering and construction industry has remained largely unchanged for decades,” said Katherine Boyle, General Partner at Andreessen Horowitz. “Unlimited’s vertically integrated, AI-first approach represents a paradigm shift, turning design and build into a rapid and continuous optimization problem. This is exactly the kind of innovation needed to restore America’s ability to build ambitious infrastructure at scale.”

The company plans to double or triple its seven-person team over the next three to six months. Most new hires will be multidisciplinary engineers who can work across mechanical, electrical, civil, and process systems using the company’s internal AI tools. The team is concentrated in San Francisco, but construction activity will continue to take place in regions where timelines are achievable. Texas is the current focus.

Modon’s goal is to build enough capacity and product maturity to compete directly with the entrenched engineering and construction incumbents that dominate industrial work. He does not think incremental tools will bend the curve. “We are not trying to give broken incentive structures better software,” he said. “We are trying to earn the right to show there is a different way to build.”