Every super has done the weekly walkthrough with a 360 degree camera bolted to the hard hat, forgotten it was recording, and later relived the moment they tripped over a conduit run while narrating to nobody. Someone in the trailer eventually watches the footage back and treats it like blooper reel material, right up until the same footage flags a missing electrical rough in before the drywall crew buries it. The joke about being filmed all day on a jobsite used to be about liability. Now it is about the fact that the footage is quietly worth something.

That weekly hard hat walk is the entire data pipeline behind Buildots, and on September 14 the company proved investors think that pipeline is worth a great deal more than a blooper reel. Buildots, the Tel Aviv founded construction intelligence startup, announced a 130 million dollar funding round led by billionaire Eyal Ofer's OG Venture Partners, with participation from Lightspeed Venture Partners, Intel Capital, Mohari Ventures, Human Capital, Qumra Capital, Avigdor Willenz, Viola Growth, and Poalim Equity. The round brings Buildots' total funding to 297 million dollars since it was founded in 2018, and lands on the back of what the company says has been threefold annual revenue growth for several years running.

The mechanics have not changed much since Buildots first pitched the idea years ago, they have just gotten more valuable. A worker wears a 360 degree hard hat mounted camera on a walkthrough that doubles as the normal quality control tour, and Buildots' computer vision and machine learning models turn that footage into a continuously updated digital representation of the project. The system compares what the camera saw against the schedule and the three dimensional building model, classifies what is actually complete at the trade level, flags deviations, and forecasts where delays are forming before they show up in a status meeting. It is reality capture that does not stop at capturing, it argues with the schedule about what it just saw.

What makes this round notable is who is cutting the checks and why. Intel Capital's continued participation, alongside a fresh lead from Ofer's OG Venture Partners, points straight at the reason Buildots says it is raising now: data centers, chip fabs, and other AI infrastructure megaprojects, the same buildout everyone from utilities to GPU makers has been racing to keep pace with all year. CEO Roy Danon framed the moment as the industry's blind spot finally getting expensive enough to fix at scale, arguing that the visibility problem Buildots solves for a data center is the same one that has quietly cost schools and hospitals for decades, just multiplied by the size of the check. The company plans to use the new capital to expand across North America and Europe, extend its platform to more stages of a project, and give owners and general contractors consolidated intelligence across an entire portfolio of jobs rather than one site at a time.

That portfolio ambition is not new talk for Buildots, it already made a move in that direction last October when it acquired Genda, a workforce and safety management platform whose clients include DPR Construction, Hensel Phelps, Clark Construction, and JE Dunn. Pairing Genda's labor and equipment tracking with Buildots' progress data was pitched at the time as an industry first: knowing not just that a task is behind schedule, but whether it is behind because of a design problem, a materials problem, or simply not enough hands on the wall that week. A funding round eight months later that explicitly talks about extending coverage "from bidding through handover" suggests that acquisition was the opening move in a longer plan to own more of the project lifecycle, not a one off bolt on.

It is worth sitting with how far reality capture has traveled from novelty to infrastructure in just a few years. A hard hat camera used to be a proof of concept demo at a trade show. Now it is the sensor layer underneath a 130 million dollar round explicitly tied to the AI industry's own physical buildout, which is a fairly circular and fairly telling detail: the models being trained in those data centers are, in a small way, funding the tool that watches the data centers get built. Whether Buildots can actually deliver portfolio wide intelligence at the scale its new backers are betting on, across hundreds of sites instead of dozens, is the harder test than any single funding announcement, and it is a test that plays out over the next several quarters of hard hat footage, not this week's press release.

Sources