Somewhere in the past few months, a bond was filed, a plot of land changed hands, and the winner of a nine-figure construction project was quietly settled.
No request for proposals had gone out. Most of the firms that might have chased the work did not yet know it existed.
That gap, between when a project starts leaving traces and when it becomes a public bid, is the wager behind Cascade, an AI pursuit platform for architecture, engineering and construction firms.
This week the company said it had raised $3.5 million from Andreessen Horowitz’s Speedrun programme, Ada Ventures, Blitzscaling Ventures, Indico Capital, shuckerVC, G2C Ventures and Snowball VC.
The premise is that a large building project sheds paperwork long before anyone issues a tender. A bond filing is a project taking shape. A property changing hands is a developer making a move.
A line item in a county capital plan is a structure that does not exist yet, and that someone will eventually win.
Cascade says it watches these signals continuously, pulling from permits, capital plans, property transactions, earnings transcripts, budget announcements and meeting minutes, and tries to connect them to where work is forming, what kind, and which firm is placed to win it.
Once it spots something, the platform scores each opportunity for fit and looks for a warm introduction through relationships already sitting in a firm’s Outlook and other software.
The comparison the founders reach for is finance. “Hedge funds look at a crazy amount of data to predict stock movements,” said Joana Ferreira, Cascade’s co-founder and chief technology officer.
“We use the same logic to predict construction projects.”
The early customer list skews toward marquee work. Cascade says its clients have touched projects including JFK and LaGuardia airports, data centres, nuclear reactors and luxury hotels, and that collectively they have surfaced more than $10 billion in project opportunities through the platform, a figure the company has not broken down.
Tim Johannesson, a principal at the luxury architecture firm Smallwood, said Cascade helped his firm find and win a $6 million project he would not otherwise have seen. “I’ve been in the industry for 30 years,” he said.
“A project broke ground a few blocks away from me. I didn’t even know it was out for bid.”
Cascade was founded by Hannia Zia and Ferreira, two former Google operators who met at UnlikelyAI, the startup set up by William Tunstall-Pedoe, the British engineer credited as a creator of Amazon’s Alexa. Zia served as vice-president of product there and Ferreira led the AI platform.
The route into construction was personal for both: Ferreira grew up in a Portuguese town built on construction and carpentry, and Zia’s father tried and failed to start a construction business of his own.
Their first customer, Munoz Engineering, whose work includes LaGuardia, Moynihan Train Hall and the Cuomo Bridge, came out of conversations Zia had with chief financial officers in New York.
“Cascade helps us plan strategically ahead in the highly competitive AEC industry in NYC,” said Syed Irfan Raza, the firm’s CFO.
Cascade’s wider bet is on the network. Even a single house pulls in roughly twenty companies, and larger projects an order of magnitude more, so the more firms that join, the company argues, the more it can match them to one another as partners, subcontractors and teaming options across regions.
Whether that compounds the way the founders hope will take longer than a few months of signings to establish.
For now the pitch is narrower and easier to test: find the work before the competition knows it is there.
Get the TNW newsletter #
Get the most important tech news in your inbox each week.