Exodigo sees what nobody else can.

Picture a Thursday morning in Brooklyn. A modest white truck rolls down a stretch of Atlantic Avenue, towing a low rig studded with antennae and sensors. No backhoe. No paint. No traffic cones strangling the block. To the deli owner watching from across the street, it looks like a vehicle that took a wrong turn off a sci-fi movie set. To National Grid, the utility company that hired it, it looks like the cheapest insurance policy money can buy.

Inside that rig, geophysical sensors are humming. Above it, in a cloud somewhere, an Exodigo model is already arguing with itself about what lies beneath the pavement. By the time the truck reaches the next intersection, a 3D picture of pipes, cables, voids and forgotten infrastructure has begun to assemble itself. When the survey ends, the verdict will be embarrassing for the city's official records: there are roughly 188% more utilities down there than the paperwork suggests. Which is another way of saying that, for decades, contractors in this neighborhood have been digging through a fog.

Exodigo's whole product is the unfogging.

Exodigo is, depending on how you squint, a sensor company, a software company, a geophysics consultancy and an AI lab. The team in Palo Alto and Tel Aviv would tell you it is, more usefully, a company that sells one thing: confidence about what is underground. That confidence is delivered as a 3D map. It is bought by transit authorities, utilities, energy giants and the kinds of engineering firms whose budgets are dictated by what a backhoe accidentally hits on day 47 of a project.

The customer list, last counted, runs past 50 organizations. Amtrak. Pacific Gas and Electric. National Grid. California High-Speed Rail. LA Metro. MTA. SEPTA. Sound Transit. In Europe, the national gas networks of Italy and France, plus a recurring credit on the UK's bottomless HS2 rail program. None of these are companies that buy enthusiastic demoware. They buy because hitting an unmapped 138 kV cable costs more than a year of Exodigo's contract.

Modern cities sit on a tangled inheritance: water mains laid by Victorians, fiber dropped in the 90s, gas lines patched in the 70s, electrical conduits that someone definitely meant to draw on a map. The honest answer to "what's down there?" has, for most of human history, been a shrug followed by a quote.

The bill for that shrug is enormous. Industry estimates put the global cost of utility strikes, redesigns and construction delays in the tens of billions a year. Every transit megaproject has a line item that quietly translates to: "money set aside for the surprise we cannot yet name." Exodigo's pitch, when you strip it down, is that this line item is a choice. Not a law of physics.

The problem is not that engineers are lazy. The problem is that the existing tools - ground-penetrating radar, electromagnetic locators, hand-drawn record drawings - each see a slice of reality and miss the rest. Asking any single one of them to map a busy urban corridor is like asking a single witness to describe a car crash from the back seat with their eyes closed.

Exodigo was founded in 2021 by Jeremy Suard, Ido Gonen and Yogev Shifman. Suard and Gonen came out of IDF Units 81 and 8200 - the technological intelligence outfits whose graduates also seeded Waze, Check Point and a long alumni list of unicorn founders. Both received the Israel Defense Prize for prior work, a credential that sounds vague in English and means rather a lot in Hebrew.

The bet was elegant. If you stop asking a single sensor to be a hero, and instead deploy many sensors at once, you get many partial truths. Run those partial truths through AI trained to reconcile them, and you get something closer to a complete answer. Multi-sensor fusion was not a new phrase - it has been a fixture of defense and geoscience for years - but applying it at construction-site scale, in software, with the speed required for real schedules, was an open problem. Suard and his co-founders bet it could be closed.

The Exodigo system is best understood in three layers. At the bottom, hardware: a vehicle-mounted, hand-pushed or drone-mounted sensor stack that combines electromagnetic, GPR, seismic and several other modalities. In the middle, the data layer: terabytes of geophysical signal funneled into the cloud. At the top, the model: AI trained to reconcile contradicting signals, simulate millions of plausible underground layouts, and converge on the one that best explains the measurements.

The output is the part the customer cares about. A 3D map, viewable in the platform or exported into the CAD and BIM tools engineers actually use. Each pipe, cable, void and anomaly comes with a confidence score, which is itself a quietly radical thing in a field that has spent a century pretending its hand-drawn record sheets were definitive.

Numbers in subsurface work are usually slippery, because nobody wants to write down what they did not know. Exodigo's are less slippery than most. In the twelve months before its Series B, the company says its surveys informed more than $75 billion of federal, state and local infrastructure investment across 18 US states, including projects with the three largest US transit agencies.

The Brooklyn job for National Grid - the one with the 188% more utilities than the records suggested - is now the case study Exodigo trots out when a skeptical procurement officer asks why the line item exists. It is a number designed to ruin dinner parties for civil engineers. It is also a number that funds an entire industry of redesigns, callbacks and lawsuits that Exodigo's customers would prefer to skip.

The investor list reads the part. Zeev Ventures and Greenfield Partners co-led the 2025 round. Returning investors 10D VC, Square Peg and JIBE re-upped. New money came in via Vintage Investment Partners and Leblon Capital. The total since 2021 stands at $214 million, with the company valued at roughly $700 million - the kind of capital stack you build when your customers are governments.

Pressed on mission, Exodigo's leadership talks about three things in roughly equal measure: safety, schedule and sustainability. Safety because most utility strikes are also human-injury events. Schedule because megaprojects live or die in the gap between drawing and dirt. Sustainability because every redesign means more concrete, more diesel, more carbon. A subsurface map sounds like a small thing. The downstream emissions of getting it wrong are not.

There is also a tidier story underneath the stated mission, which is that Exodigo's founders spent their formative years in intelligence units whose entire job was to draw confident maps of partly-known reality. That habit doesn't really leave a person. It just changes targets - in this case, from cross-border tunnels to subway alignments and gas networks.

Three trends bend the demand curve in Exodigo's direction at once. First, the global infrastructure backlog: aging utility networks in the US and Europe, plus a transit build-out in Asia and the Gulf that has no historical precedent. Second, the rise of geotechnical risk as a board-level concern, especially as climate-driven flooding and subsidence start to reshape soil under existing roads and rails. Third, the maturity of AI models that can now make sense of messy, multimodal sensor data without an army of geophysicists post-processing every signal.

Exodigo is not unique in noticing this. Traditional subsurface utility engineering firms have decades of relationships and reputation. Standalone GPR specialists exist in every city. Geospatial AI startups are circling adjacent problems. What Exodigo has - so far - is the rare combination of multi-sensor hardware, real models, real revenue, and a customer base that already trusts it with projects measured in billions. The moat is unsexy. It is also exactly the kind of moat that holds.

By late afternoon, the white truck that started this story is gone. The deli owner has stopped paying attention. The block looks exactly as it did before - no excavation, no spray paint, no traffic cones. The only thing that has changed is the file sitting in a project manager's inbox at National Grid. In it, the underground has been re-drawn. The 188% gap between records and reality is closed. The redesigns that would have happened in month nine of construction will not happen now. The lawsuit that lives inside every unmapped cable just lost its address.

That is what Exodigo sells. Not maps, exactly. The quiet erasure of surprises. In an industry that has spent a hundred years pretending it knew what was down there, the most radical product turns out to be the one that actually does.