Eric Wu built and ran Opendoor, one of the most ambitious real estate startups of the last decade, before retiring in 2022 after rapidly rising interest rates sharply slowed home sales. He spent a year rebooting (he had been running the company for eight years at that point) and could have easily jumped into investing when he felt his hiatus was over. But like many founders, he’s so convinced that AI will be the defining technology platform of his life that he more recently decided to dive back into building companies. As he told me during a call earlier this summer, “I knew that if I looked back in 10 years and didn’t do something related to it, I would probably regret it.”
The opportunity Wu is pursuing—building AI co-pilots for construction workers and other field workers, something he describes as a hands-free expert trainer for people who physically build things—isn’t exactly a secret. The construction industry is short of workers by a wide margin.
The trade group Associated Builders and Contractors has said approximately 349,000 additional workers It would take this year alone to keep pace with construction demand, a challenge that appears to get worse with workers. getting olderworkers get shipped outside the United States due to more aggressive immigration enforcement in the United States, and also the increasing number of major projects being planned, specifically data centers.
We’ve all heard by now, from across the AI industry, about the gigantic server farms that the rise of AI demands. These projects have grown enormously and the demand for personnel has grown with them. While a large data center campus once needed around 750 workers at its peak, larger projects now require much more. Meta’s Hyperion campus in Richland Parish, Louisiana, for example, will require approximately 5,000 construction workersand OpenAI’s Stargate site in Abilene, Texas, have reported that they are involved 6,400 workers.
Kelly, the global staffing company, has said that 90% of data center operators They now cite staff shortages as a critical limitation to their ability to build or expand.
Wu’s new company, called NavigateAIaimed at addressing that layer of people, officially launched in late May with $25 million in seed funding and a $225 million post-money valuation led by Elad Gil, with participation from Khosla Ventures, Fifth Wall, real estate giant Lennar, Tishman Speyer, electrical contractor Helix Electric and a roster of angels including DoorDash’s Tony Xu, Instacart’s Apoorva Mehta and Coinbase CEO Brian Armstrong.
None of those backers are particularly surprising, given Wu’s connections to the real estate industry, as well as Silicon Valley biggies. Gil, for example, was an investor in Opendoor and Keith Rabois of Khosla Ventures co-founded Opendoor. Khosla Ventures founder Vinod Khosla has also been outspoken in recent years about the dozens of startups in the company’s portfolio that are being built around AI “workers” of various types, whether they be oncologists, chip designers or construction workers.
Wu’s core product runs on smartphones and, hands-free, through Meta’s AI glasses. The idea is for a construction worker to point the camera at what they are building and ask, in plain language, if it is installed correctly, if the torque is correct, if it meets code, etc. Wu says NavigateAI can query construction specifications, manufacturer manuals, and company policies in real time. He also says that the hands-free experience is considerably superior (you won’t want to look at a phone if you can help it) and that, in fact, the company is working with Meta to get the glasses safety certified for environments where protective glasses are required.
NavigateAI is also partnering with AIM, a Meta-backed fiber installation career school that guarantees job placement to graduates, giving the company a channel to reach workers before they set foot on a jobsite.
Getting workers comfortable with AI-assisted work during training is apparently very important. Wu says adoption curves split sharply between younger workers, who he says embrace the product, and 30-year-old journeymen, who trust their own instincts and aren’t exactly lining up to strap a computer to their faces.
In terms of business model, the company started with a token-plus-margin pricing structure, similar to usage-based SaaS, but has migrated its newer contracts to a share of the value created. For example, if NavigateAI helps a builder reduce the total cost of a home from $300,000 to $280,000, the company gets about 20% of those $20,000 savings. Wu told me that Lennar—one of the nation’s largest home building companies and another of NavigateAI’s investors—spends about $9 billion a year on labor, installation and construction, so even an improvement in 5% to 10% would represent hundreds of millions of dollars in potential value.
The long-term strategy, which Wu is candid about, is data. Every job completed with NavigateAI generates tagged egocentric videos of field workers building and maintaining physical things correctly and incorrectly, and that’s a data set that Wu believes will eventually be worth as much to robotics companies as Navigate’s own software business is to construction clients.
Of course, there are real challenges. Value-based pricing can create an attribution dilemma, as Wu openly acknowledges. Proving that a house was built faster because of NavigateAI and not because of the sunny weather, or the particular team at work, or the availability of materials, requires A/B testing across divisions and is approximate, by his own admission. You could imagine that a customer’s dispute over savings attribution gets complicated quickly, even if that customer is also an investor in NavigateAI.
That resistance from experienced workers also seems to be a pretty serious problem. The journeyman veteran is whose industry knowledge would make NavigateAI’s product smarter, and no trade school can fully replace that knowledge. Then there is the responsibility of safety. If the NavigateAI software deletes a connection that then fails, what happens? Defect liability is a notoriously litigious area and it is not yet clear how these issues will be handled as they invariably arise.
There is also, as always, the competitive issue. When we talk, Wu mentions that the most common current alternative is for a worker to search something on Google or ask on ChatGPT, and that NavigateAI can go much further than that. But all the big LLM companies have model capabilities and, in the case of Meta, also hardware distribution. As unlikely as it may be, they could create their own businesses.
Of course, the most likely threat is a similar player. NavigateAI’s defense relies on proprietary data and workflow integrations that take years to accumulate. Wu himself mentions Buildots and OpenSpace as the closest points of comparison, but says they are “more focused on project management,” while Navigate is based on “individual work.”
Either way, in Silicon Valley, the network sometimes matters as much as the product or competitive landscape, and the combination of Wu, Gil, Khosla Ventures and Lennar, among others, serves as a strong signal. Furthermore, Wu doesn’t seem willing to worry about competitors right now. He just seems excited to be building something new, at a time when not building it would seem like a mistake.
Underscoring that point is that it doesn’t yet have a board of directors, and says it will “try to go as long as possible without one,” so it can focus on customers rather than governance. A line can be drawn between that observation and his time leading Opendoor, which went public through a special purpose acquisition company in late 2020. Being CEO of a public company made the parts of the job he wanted to spend his time on more difficult to do, forcing a constant balance between building products and managing a board of directors and shareholders. For now, it’s clearly a trade-off not to be missed.
When you purchase through links in our articles, we may earn a small commission. This does not affect our editorial independence.