An architect in Brooklyn walks into a 1920s brownstone with a tape measure she will not need. She pulls out an iPhone, taps a green dot on a dashboard called Polycam, and starts walking. Six minutes later, the building exists as a textured 3D model inside her browser. Tomorrow it will live inside her CAD software. Yesterday, a job like this took two technicians and three days.
This is the unglamorous, mostly invisible work that Polycam is in the middle of - the kind of shift that does not announce itself. It just turns up in workflows. A retail chain measures every storefront in a quarter. A film crew scans a back-alley in Brooklyn for a fight scene. A father in Oslo scans his daughter's clay dragon before she breaks it. None of them are thinking about LiDAR sensors or photogrammetry. They are thinking about getting on with the next thing.
Polycam, the Berkeley company that built the thing, has so far raised $22 million, employs around 58 people, and counts Adobe, Sequoia, and YouTube co-founder Chad Hurley among its backers. It also serves as the spatial capture layer for CVS, Wayfair, Warner Bros Discovery, Amazon, and Exxon Mobil. The list is more eclectic than impressive, which is the point.
The Problem Nobody Asked To Solve
For most of the last fifty years, 3D capture was an enterprise sport. You hired a survey team. You leased a $40,000 laser scanner the size of a microwave. You waited a week for someone in another time zone to clean the point cloud. The output went into a CAD file that maybe twelve people in the building knew how to open.
This worked in the sense that it kept working. It also meant that anything below the multi-hundred-thousand-dollar budget line had to be measured by hand, sketched, or simply guessed at. A staggering amount of construction, real estate, insurance, retail and forensic work still leans on that last category.
The other inconvenient truth: the rest of imaging long ago left the lab. Photography became casual in 1900. Video became casual in 2007. 3D capture, stubbornly, did not. Until iPhones started shipping with LiDAR sensors in late 2020, nobody had a real reason to fix that.
The Founders placed an awkward bet
Chris Heinrich and Elliott Spelman met at Ubiquity6, a mobile-3D startup later folded into Discord. They were the kind of pair you get when a physics PhD and a product builder share an office for too long: convinced that the technology was almost ready, that the timing was the only thing left to argue about.
When the iPhone 12 Pro shipped with a LiDAR sensor, they decided the timing argument was over. Polycam launched in late 2020 as an app store experiment. In its first full year it made $280,000 - the kind of number that gets you a meeting in Berkeley and a polite no in Sand Hill Road.
The bet was simple, and slightly heretical: 3D capture would not be unlocked by better hardware. It would be unlocked by better software that hid the hardware. You should not have to know what photogrammetry is, in the same way you do not have to know what an autofocus algorithm is when you open the camera app. Polycam decided to be the camera app.
// Milestones
The Product is six tools in a trench coat
What started as a LiDAR scanning app has, in five years, become something closer to a Swiss Army knife for the physical world. Polycam now ships seven distinct capture and editing tools, each one ordinary on its own and faintly strange when stacked together.
Spatial Capture
Scan rooms, sites and entire buildings with sub-centimeter accuracy from a phone in your hand.
Object Capture
Photogrammetry and Gaussian splatting. Walk around an object. Get a high-fidelity 3D model.
Floor Plans
Automated 2D and 3D plan generation. Useful for realtors. Annoying for tape-measure salesmen.
Drone Mapping
Aerial photogrammetry for sites, roofs and outdoor environments.
Apple Vision Pro
Spatial capture for the headset that wants to be a workstation when it grows up.
Polycam Web
Editing, sharing, asset libraries and CAD/BIM export in a browser.
Object Capture API
The same engine, available to developers who want 3D inside their own apps.
The Proof is in the receipts
It is one thing to ship a clever piece of software. It is another to be invoiced by Exxon for it. Polycam has been doing both. Revenue went from $280K in 2021 to $1.8M in 2022 to $6.5M in 2023 - a 23x climb that does not happen by accident, and almost never happens in 3D tooling. Most of that came from the long tail: solo architects, indie developers, small surveying outfits. The fact that Fortune 100s followed is the part that got Silicon Valley's attention.
Polycam Annual Revenue
The Mission is unfashionably literal
Polycam's stated mission is to "democratize 3D creation." It is the kind of phrase that has been worn smooth by overuse, and which the company has the inconvenient habit of meaning. The free tier of the app does most of what the paid tier does. The web library is open by default. The Object Capture API is published. None of this is normal for a startup that just took a Series A.
Heinrich's Kodak Brownie line is the cleanest articulation of the bet. The Brownie, in 1900, was the first camera ordinary people could afford. It did not invent photography. It made photography a thing households had. Polycam thinks 3D is at the same moment: the technology is essentially solved, the hardware is in pockets, the only thing left to do is shrink the friction to zero and let the rest happen.
Whether that produces millions of casual home scans, or quietly replaces the survey industry, or both, is less interesting than the fact that Polycam clearly believes either one is fine.
Why It Matters tomorrow
The next layer up from spatial capture is spatial intelligence. Once a building exists as a model, you can run analytics on it. Once a city block exists as a model, you can simulate flood scenarios. Once a crime scene exists as a model, you can revisit it from the jury box. The capture is the unsexy plumbing; everything interesting happens after.
Polycam's 2025 enterprise platform - collaborative, organization-wide, with audit-grade exports - is a quiet move in that direction. Gaussian splatting, which was barely a research paper two years ago, is now a button in the app. The Apple Vision Pro version exists less because the headset is popular and more because Polycam wants to be there when it might be.
The thing about reality capture, as a category, is that it is dull right up until it isn't. The companies that get there first tend to look like overnight successes the morning after a decade of quietly stacking the deck. Polycam, five years in, is doing the stacking.
Back to the brownstone in Brooklyn. The architect has uploaded her scan, exported it to her CAD software, dropped in a load-bearing wall and shared a link with her client. It is 4:47 p.m. She is leaving early.
This is, in a way, the entire story. The tape measure is in a drawer somewhere. The survey team she didn't hire is still busy with someone else. The model on her screen, which would have cost five thousand dollars and a week six years ago, cost her one walk through a house and an iPhone she was already carrying.
Polycam is not glamorous, exactly. It is not trying to be. It is just the green button between the world as it is and the world as it gets used in software - and there is more of the world left to capture than anyone has bothered to count.