Loft Orbital. Space, made simple.

A satellite leaves Vandenberg. Nobody at Loft built it for one customer.

It is a Thursday in early 2026. A SpaceX Falcon 9 lifts off the California coast. Tucked inside the fairing is a 100-kilogram satellite called YAM. Yet Another Mission. On board: a hyperspectral imager from one customer, a software-defined radio from another, an AI inference workload from a third, plus a couple of guest apps that didn't exist when the spacecraft was being assembled. None of these customers own the satellite. Most of them have never touched one.

That is the new normal Loft Orbital is building. The company sells space the way Amazon sells servers. You don't buy the box. You pay for what runs on it.

Building a satellite is the easy part. Everything else is the trap.

For decades, the path to orbit was a private endurance test. Pick a vendor. Sign a long contract. Design a bespoke spacecraft. Wait three to five years. Hope the launch window holds. Operate it yourself, twenty-four hours a day, with a team you have to recruit. By the time the data starts flowing, the original use case has half-evaporated.

The economics worked when the customer was a national government. They struggled when the customer was an AI startup with a sensor it wanted to try in LEO for eighteen months. There was no middle path - until someone built one.

Three Spire alumni made a small, slightly heretical wager.

Pierre-Damien Vaujour, Alex Greenberg, and Antoine de Chassy met at Spire Global, a satellite-data company where they'd seen the workflow from the inside. In 2017 they walked out and started Loft Orbital with a single counterintuitive idea: most customers don't want a satellite. They want a result. So build one standardized bus, design a universal payload hub, and rent floors of it.

The investors who got it first were small. Uncork Capital and Foundation Capital led a $3.2M seed in 2018 - a round so modest that SpaceNews described the pitch with the word "condo." The headline almost wrote itself. Condosats. A multi-tenant building, two hundred miles up.

YAM, Cockpit, and the universal port for hardware in orbit.

The fleet has a name only a developer could love: YAM, short for Yet Another Mission. Each YAM satellite is a roughly 100-kilogram bus paired with Loft's payload hub - a standardized interface that handles power, data, mechanical mounting, and thermals so a customer's hardware can just plug in. On the ground, a mission operations platform called Cockpit hides all of the spacecraft's complexity behind a single screen.

Then in 2023 Loft added the part that actually surprised people. Virtual Missions. Customers can now deploy software apps to Loft's in-orbit compute, including AI workloads running on GPUs. YAM-6 was the first satellite built for this. Space Compass committed to running an AI-based ship detection app on it. Helsing followed with onboard radio-frequency intelligence. The satellite stopped being a thing you sent up. It became a runtime.

What you actually get from Loft

How a "condo" pitch turned into a $500M backlog.

The bookings number is the one that quiets the room. Per TechCrunch, Loft had logged more than $500 million in customer contracts on roughly $160 million of capital before the Series C even closed. That's an unusual ratio in space, where companies are usually heavy on outflows and shy about backlog.

Make space feel like cloud. Not metaphorically. Operationally.

The Loft Orbital pitch is not "space is exciting." It's that space is a layer of infrastructure, and infrastructure should be boring in the same useful way that a database is boring. Predictable interfaces. Standardized hardware. A control plane that abstracts complexity. A pricing model where you pay for what you use and skip what you don't need.

For a customer, that re-shapes the question. The old one was "can we afford to build a satellite to test this idea?" The new one is "can we ship the payload before the next budget cycle?" The first question kills projects. The second one starts them.

Software is moving into orbit. Loft is laying the rails.

The next decade in space is unlikely to be about more rockets. It will be about what runs on the satellites that are already up there. AI inference at the sensor. Hyperspectral imaging fused with synthetic-aperture radar. Edge processing that compresses 50 gigabytes of data into a 5-kilobyte alert before it ever touches a downlink. Quantum payload experiments that need a few hundred kilograms and a polite operator.

None of that is helped by a model where every team has to build, launch, and operate its own spacecraft. All of it is helped by a layer that says: bring the workload, we'll handle the rest. That layer doesn't exist yet at the scale anyone needs. Loft is the company most credibly trying to build it.

The skeptic's response is fair. Plenty of companies have promised to be the AWS of something. Most of those somethings did not get an AWS. But the receipts here are unusual. The customers are not theoretical. The satellites are flying. The bookings are large enough to be embarrassing if they don't ship.

Same launch. Different industry.

Return to the Falcon 9. Same coast, same weather, same fairing. What's changed is the manifest inside. The hyperspectral imager belongs to a startup that didn't exist five years ago. The radio belongs to a defense AI company. The compute is rented by the hour. Two of the apps onboard were pushed in a software update last month - after the satellite was already in orbit.

None of those customers had to build a satellite. None of them had to hire a mission ops team. None of them had to wait four years to find out if their idea worked. That is the change Loft Orbital has been pushing on, quietly, for about eight years now.

Space, made simple. The tagline is doing a lot of work. So far, the company is too.