Encord Secures $60M to Fuel Physical AI Data Infrastructure

Jan de Vries · 2026-02-27

Let's talk about something that's easy to overlook when we get excited about robots and self-driving cars. It's not the flashy models or the powerful chips. It's the data. And that's exactly where London-based Encord just got a massive $60 million vote of confidence. This Series C funding round, led by Wellington Management, pushes their total funding to a cool $110 million. It's a serious bet on their vision as physical AI systems finally move out of the lab and into the real world. ### The Real Bottleneck for Physical AI Everyone's racing to build bigger, more complex AI models. But Encord's co-founder, Ulrik Stig Hansen, puts his finger on the real issue. "For physical AI, the bottleneck isn't model size," he says. "It's data readiness." Think about it. You could have the most brilliant AI brain ever designed. If you feed it messy, inconsistent, or unrealistic data, it's going to fail spectacularly when it tries to navigate a busy street or operate a robotic arm. Encord's whole purpose is to solve that foundational problem.  ### A Wave of European AI Investment This funding didn't happen in a vacuum. There's a tidal wave of capital flowing into Europe's AI scene right now. Just look at the numbers from the last couple years: - In France, Mistral AI secured a staggering $2 billion for model development. - The UK's Nscale raised over $1.1 billion to expand its AI cloud infrastructure. - Berlin's Helsing closed a $700 million round for its defense software platform. Combined, we're talking about over $3.5 billion in disclosed funding. It's a clear signal that investors are all-in on the future of AI, from the models themselves to the hardware that runs them. Encord's raise is a crucial piece of that puzzle. While others focus on the models and the computers, Encord is building the indispensable data layer that makes everything else work. ### Why the Data Layer is a Game-Changer Bill Tinney from Encord customer Vantor puts it perfectly. "For production AI teams, how you operationalize your data is a core competitive advantage." At Vantor, which works on AI for critical infrastructure and national security, they needed a platform that could handle immense complexity without falling apart. Encord provides a unified system that manages the entire AI data lifecycle—curation, annotation, evaluation—all in one place. No more juggling a dozen different tools that don't talk to each other. Founded in 2021, Encord acts as a universal data layer for AI teams. They're already working with over 300 companies, including big names like Woven by Toyota and Zipline. This new cash injection will help them scale their platform to handle the wild, multimodal data that physical AI craves. We're talking about audio, video, sensor streams, 3D point clouds—formats that give old-school data platforms a serious headache. ### The Physical AI Revolution is Here Here's the big picture. After years of promises and pilot programs, physical AI is hitting its stride. Analysts project a mind-blowing 400 million AI-powered robots coming online in just the next four years. The entire industry is poised to blast past $30 billion in that same short window. This isn't like the large language models that learned from the public internet. Physical AI—the brains in robots, drones, and autonomous vehicles—learns from proprietary, real-world data. Sensor feeds, video from the field, robotic telemetry. Storing and making sense of that data is a monumental challenge, and it requires a new kind of infrastructure. That's the gap Encord is filling. As these systems transition from fascinating demos to essential parts of our daily lives, the company ensuring their data is ready, reliable, and real-world aligned isn't just helpful. It's absolutely critical. This $60 million round is a bet that they're the ones to build that essential foundation.