Outsight.

Who they are, this minute

A Boeing 777 is parking at Gate D32 at Dallas Fort Worth. Three hundred passengers are about to flood a corridor sized for two-fifty. Nobody has spotted a camera. Nobody needs to.

Above the jet bridge, a small box is quietly counting bodies as shapes - tall ones, short ones, ones pulling carry-on. The count lands in an airport operations dashboard before the first passenger has reached the people-mover. The box is a LiDAR sensor. The brain inside it is software from a Paris company called Outsight. The whole conversation happens in milliseconds, and at no point does anyone's face enter the room.

That, in a sentence, is the Outsight pitch: the physical world made legible to machines, without making the people inside it identifiable.

Why nobody could read a 3D point cloud

LiDAR is the sensor that makes self-driving cars not crash. It throws out millions of laser pulses a second, builds a 3D map of whatever is in front of it, and updates ten or twenty times a second. The hardware industry exploded after the DARPA challenges. The software industry, less so.

What you got, in 2018, if you bought a LiDAR was: a firehose of points. A grey snowstorm of dots, refreshing endlessly, beautiful and entirely useless unless you happened to employ a team of computer-vision PhDs to write everything from raw data down to "that is a forklift."

Hardware makers shipped 50 different brands of sensor, each with its own driver, its own coordinate system, its own quirks. Operations teams - the people who actually run airports and train stations - had no chance.

The industry called it "the integration problem." Outsight's founders called it "a Tuesday."

Four founders, one stubborn idea

In July 2019 four engineers got together in Paris and decided to fix the layer that nobody else wanted to own. Raul Bravo and Olivier Garcia came from Dibotics, a startup that had been quietly impressive at simultaneous 3D localisation and mapping. Scott Buchter came from Lasersec. And Cédric Hutchings - the one investors remember - had co-founded Withings, the connected-health company Nokia paid €170 million for.

Hutchings could have done anything next. He went looking for the next physical-world platform: not for wrists, but for spaces. The bet was simple enough to fit on a napkin and ambitious enough to take a decade: build the operating system that sits between any 3D sensor and any operational team, regardless of brand, regardless of use case.

The wager, in plainer English

Be the Switzerland of LiDAR software. Don't sell sensors. Don't pick favourites. Just be the layer that turns whatever-someone-bought into "here are 1,247 people in Terminal 4, here is the queue at security, here is the curbside truck blocking lane two."

It is the kind of bet that sounds obvious in retrospect and was anything but obvious in 2019, when most LiDAR companies were busy promising autonomous cars by Tuesday.

What the software actually does

Strip away the deck slides and Outsight does three unglamorous things very well.

One, it ingests. Point in any major-brand LiDAR sensor - Ouster, Hesai, Innoviz, Velodyne, you name it - and Outsight will speak to it. No bespoke driver to write.

Two, it understands. The platform classifies what is a person, what is a vehicle, what is luggage, what is a forklift, and tracks every one of them through space and time at 10-20 frames a second.

Three, it produces KPIs. Queue length. Dwell time. Curb occupancy. Tarmac intrusion. Vulnerable-road-user near-miss. All anonymous. All real time. All deliverable as alerts, dashboards, or REST APIs your existing systems can swallow.

A privacy posture that wasn't an afterthought

The interesting thing about LiDAR, the one cameras can never offer, is that it cannot see a face. It sees a moving cylinder roughly six feet tall. That single property turned out to be a regulatory cheat code in Europe, where GDPR rules around camera-based crowd analytics are - to use the technical term - terrifying.

Customers who paid, not just clapped

Outsight's product pages are not the proof. The proof is the boarding pass in your back pocket. Dallas Fort Worth, one of the busiest airports in the world, did not buy the largest 3D LiDAR network in the industry because it was charming. It bought it because the math on passenger-flow optimisation pencils out fast when you can save thirty seconds at security and multiply by 75 million annual passengers.

Aeroporti di Roma started with a pilot in the arrivals immigration hall at Fiumicino, liked the numbers, and is now extending Outsight across nearly every Schengen common-use area in the airport. Groupe ADP - the Aéroports de Paris group - is both a customer and a strategic investor, which is the kind of vote of confidence money cannot quite buy on its own.

Then there is the partner column. Intel signed a strategic collaboration in April 2026 to push Physical AI deeper into the enterprise edge. The major LiDAR sensor makers - Ouster, Hesai and the rest - are partners not by accident: a multi-vendor software player makes their hardware easier to sell. Everyone wins. Quietly.

An invisible nervous system for the world

Ask any of the founders what Outsight is really building and you get a version of the same answer: a nervous system for places. The cells are sensors. The neurons are point clouds. The cortex is software that turns motion into meaning.

It is a mission that gets less ridiculous every year, mostly because the alternative - a future of physical spaces instrumented by cameras with face recognition - is the version nobody outside of a sales deck actually wants. Spatial intelligence done right is the version of ubiquitous computing that does not require ubiquitous surveillance.

Why investors keep showing up

Outsight has raised roughly $44.2 million across two rounds. The cap table is unusual: pure financial investors (Energy Innovation Capital, Demeter), strategic operators (Groupe ADP, Safran, Faurecia, BNP Paribas) and the French state's defence-innovation arm (Bpifrance). It is the sort of mix you see when the technology is dual-use and the customers are critical infrastructure.

Why this matters in six years

The Louvre heist, in late 2025, made an embarrassing point that the security industry had been politely avoiding. Cameras are excellent at recording crime and only adequate at preventing it. LiDAR networks, properly tuned, are something closer to the inverse: they raise the alarm before the glass case opens. LiDAR Magazine ran a long piece in October 2025 about how Outsight's stack could have rewritten that night.

Multiply that by every museum, every airport perimeter, every tarmac, every six-lane intersection where a cyclist is about to be invisible to a turning truck. The market for spatial intelligence is not a niche. It is the rest of the internet of things, finally with a brain attached.

That is also why the Intel partnership matters. Edge compute is the bridge between "LiDAR data is interesting" and "LiDAR data ships with the building." When that bridge is built, Outsight is the obvious software companion - in the same way that a thousand SaaS companies once rode the cloud-compute wave.

Back to the gate

The 777 has finished parking. The corridor at Gate D32 is full. The operations team has had the heads-up for ninety seconds and has already opened a second lane at the people-mover. Three hundred passengers will be at baggage claim seven minutes ahead of schedule. None of them saw a camera, and none of them needed to. The sensor on the jet bridge has already forgotten them, because it never knew their names in the first place.

That is what Outsight does. Quietly, repeatedly, in a hundred buildings most of us will pass through this year without thinking twice. Which is, on reflection, the very nicest review a piece of infrastructure software can get.