Noetive emerged from stealth on 16 September with a $41 million seed round led by Eclipse, announced from San Francisco. The company describes itself as an “AI NeoLab” building what it calls an intelligence of record for the physical economy: agentic software that learns how an industrial business runs, paired with the company’s own sensing hardware.
Construction is one of three verticals named on the company’s site, and the release places it among the sectors where Noetive says it already operates: the boilerplate describes the company as working “through real-world deployments across manufacturing, logistics, construction, and other industrial sectors.” It is also the vertical with no customer named in the announcement, and both halves of that are why the round is worth covering here.
The Round
Alongside Eclipse, the release names Craft Ventures, The Westly Group, Swish Ventures, Factory, Incite Ventures, Gigascale Capital, Operator Partners and Liquid 2 Ventures.
The angel list is unusually heavy for a seed, and per the announcement includes Meta CTO Andrew “Boz” Bosworth; Airbnb CTO and former Meta VP of GenAI Ahmad Al-Dahle; Nest co-founder and Incite Ventures founder Matt Rogers; Carnegie Mellon professor and founding director of Meta Reality Labs Pittsburgh Yaser Sheikh; Gigascale Capital founder Mike Schroepfer; former Founders Fund and Goldman Sachs investor Zach Frenkel; Decart AI co-founders Dean Leitersdorf, Moshe Shalev and Orian Leitersdorf; and Boris Sofman, co-founder and CEO of Bedrock Robotics — whose company raised $270 million for job-site autonomy in June and put autonomous excavators on live sites in August.
Eclipse did not merely lead the round; its founder and CEO Lior Susan says the firm built the company. “Company building has always been core to how we operate at Eclipse,” he said in the announcement. “AI has transformed the digital world, but the $30 trillion physical economy has largely been left behind. We saw an enormous opportunity to change that, and built Noetive to go after it.” That is a venture-studio origin rather than a founder-led one, and it is stated openly rather than buried.
The connection is closer still: the release says Eclipse partnered with “Frenkel, the firm’s Chief AI Officer, from the earliest stages of company creation.” The chief executive was Eclipse’s own AI chief before he was Noetive’s founder.
Chief executive and co-founder Amir Frenkel spent close to a decade as a VP at Meta and held leadership roles at Alphabet and Amazon, according to the release. His framing of the gap:
“AI will transform and reimagine the potential of companies in the physical economy, but current AI solutions were designed for information work living on the internet. Unlocking that potential requires intelligence that learns alongside the operators running the physical business, coordinating people, machines, and digital systems so every decision makes the entire operation smarter. Noetive is an AI NeoLab for operational technology to bring that intelligence into the real world.”
The Construction Pitch
Noetive’s product has two parts: agentic software that sits on top of the systems an operator already runs, and a universal sensing pod that, in the company’s words on its own site, “provides spatial-intelligence and closes the loop between planning and actual operations.”
The construction section of that site is headed “Closing the plan vs site gap,” and reads:
“No plan survives the job site. Schedules are built in the office and start drifting the day work begins. The real status lives in a superintendent’s head and a stream of photos and phone calls that reach the plan days late, if at all. We bring sensing to the site and close the loop between the plan and what’s actually happening on the ground, so slippage shows up as it starts, decisions are made on today’s reality instead of last week’s, and the whole project stays coordinated.”
Whoever wrote that has been on a job site, or listened carefully to someone who has. “The real status lives in a superintendent’s head” is the actual condition, described without the usual vendor euphemism about digital transformation.
The strategic choice underneath it is the sensing pod. Shipping your own hardware into an industry that already has cameras on site is a decision to own the data capture rather than integrate with whatever is there — an expensive path, and the one that makes the “self-improving” loop possible if it works.
What Is Actually Established, and What Is Not
The only customer named in the announcement is Steuben Foods, a food manufacturer, whose CEO Menachem Katz describes a production-planning deployment: “Work that once happened monthly and took a week of planning now happens daily and takes minutes, allowing us to respond to new orders and changing market conditions in real time.”
That is a specific, credible account — of a factory. It says nothing about a job site, and we should not let it. The construction vertical, as of this announcement, is a described capability and a stated market, not a deployment we can name.
The release also says Noetive “is already working with strategic design partners proving out the value and applicability of its technology in industries including manufacturing, logistics, energy, and data centers.” Construction is not in that list, though it does appear in the company boilerplate quoted earlier. Both sentences are in the same document; we are reporting both rather than picking the one that suits the story.
We also do not know: how many sensing pods exist, whether any are installed on a construction project, what the product costs, what the headcount is, or what the company’s total capital raised amounts to beyond this round. None of that is in the release and we did not find it elsewhere.
The Category It Is Walking Into
Analysis. Construction progress tracking is not an empty market, and $41 million of seed capital buys entry rather than position. Buildots raised $130 million on exactly this problem — comparing captured site reality against the schedule and the model — and did it with a customer list rather than a thesis. OpenSpace has documented 1,000 data-center projects. Both companies have years of captured sites behind them.
Noetive’s differentiation, if it has one, is not the tracking. It is the claim that the same intelligence layer spans a factory, a warehouse and a job site, and gets better at each because it has seen the others. That is a real hypothesis and an unproven one. The counter-argument is the oldest in enterprise software: the operator who needs schedule slippage detected does not care whether the model also plans pallet consolidation, and the vertical specialist usually wins the account.
There is also a reading in which the construction vertical is a demand signal rather than a product bet. The people who put money in — Sofman most pointedly, running a construction-autonomy company of his own — are not naive about how hard job sites are. That Caterpillar’s physical-AI deal is about watching rather than driving is the same instinct from the opposite end of the market: perception first, action later. Sensing the site is the tractable half.
The Honest Caveats
Noetive did not name a construction customer in this announcement, and we did not find one on its site or in the coverage of the round that we checked. Everything in the construction section above is drawn from the company’s own website and release. We are reporting a positioning, and we have labelled it as one.
The Steuben Foods outcome is the customer’s own characterisation, reported as given. We have not seen the underlying measurement.
“Emerged from stealth” means the record is thin. There is no published product history and no pricing, no published construction deployment, and no independent account of the technology working — the one customer account we have, from Steuben Foods, is the customer’s own and describes a factory. A large seed from credible investors is evidence about the investors’ judgement, not about the product.
The angel titles are as the release states them. We confirmed Boris Sofman’s role at Bedrock Robotics against our own prior reporting; the remaining titles are reproduced from the announcement and we did not separately verify each one.
“AI NeoLab” is the company’s term, not a category. We have used it in quotation marks throughout because it describes how Noetive wishes to be understood rather than anything a reader can check.
What to Watch
The first named construction deployment, and how long it takes. A company that ships its own sensor hardware has a much slower proof cycle than one selling software against existing cameras — pods have to be manufactured, installed, and survive a site.
If a named contractor is talking about Noetive on a real project within a year, the multi-vertical thesis has something behind it. If the construction page is still a page in twelve months while the factory business compounds, then this is an industrial-AI company that lists construction, and the distinction will matter to anyone deciding whether to take the meeting.
