Encord, the info infrastructure firm for bodily AI, immediately introduced a €50 million ($60 million) Sequence C in an effort to speed up product improvement, develop into new markets, and scale its AI-native information infrastructure platform as bodily AI methods transfer from pilot programmes into full manufacturing.
The spherical was led by Wellington Administration, bringing the corporate’s complete funding to €93 million ($110 million). Present traders Y Combinator, CRV, N47, Crane Enterprise Companions and Harpoon Ventures additionally participated within the spherical alongside new traders Brilliant Pixel Capital and Isomer Capital.
“Everybody is concentrated on constructing greater fashions,” says Ulrik Stig Hansen, co-founder and co-CEO of Encord. “However for bodily AI, the bottleneck isn’t mannequin measurement. It’s information readiness. You’ll be able to have probably the most subtle mannequin on this planet, and it’ll nonetheless fail if the info feeding it’s incomplete, inconsistent, or misaligned with real-world situations. That’s the issue we clear up.”
Encord’s €50 million Sequence C comes amid sustained capital flows into Europe’s AI infrastructure and autonomy stack in 2025–26.
In France, Mistral AI secured €1.7 billion in Sequence C funding to advance mannequin improvement and develop cloud infrastructure for enterprise deployment. Within the UK, Nscale raised €958 million in Sequence B financing to develop its AI cloud and information centre footprint, whereas London-based Stanhope AI attracted €6.7 million in Seed funding to develop adaptive AI methods for robotics and defence functions. In the meantime, Berlin’s Helsing closed a €600 million Sequence D to develop its AI-enabled defence software program platform.
Mixed, these rounds characterize over €3 billion in disclosed funding, underlining the size of funding directed in the direction of AI fashions, compute infrastructure and physical-world functions.
In opposition to this backdrop, Encord’s elevate displays continued investor deal with the info layer underpinning bodily AI methods, complementing bigger financing rounds concentrating on mannequin improvement and compute capability.
Invoice Tinney, Senior Director of AI Product Administration and Partnerships at Vantor, an Encord buyer, provides, “At Vantor, we construct AI for essential infrastructure and nationwide safety – we would have liked an information platform that might match our ambitions. Encord provides us a unified information layer that scales with the complexity of our geospatial workflows, from curation to annotation to analysis, with out device fragmentation. For manufacturing AI groups, the way you operationalize your information is a core aggressive benefit.”
Based in 2021, Encord is a common information layer for AI. The platform helps AI groups prepare and run their fashions with the correct information – managing, curating, annotating, and aligning information throughout the total AI lifecycle.
Encord works with over 300 AI groups, together with Woven by Toyota, Zipline, AXA, and Skydio.
As we speak’s funding will assist Encord scale its AI-native information infrastructure platform, which helps AI groups handle, curate, annotate, and align the multimodal information that bodily AI methods rely on, together with audio, video, photographs, sensor information, 3D level clouds and different codecs that legacy information platforms have issue dealing with.
The corporate says that their Sequence C comes as bodily AI – which powers robots, autonomous autos, drones, and different methods that function in the actual world – enters an new progress stage. After years of lab demos and pilot programmes, these methods are shifting into manufacturing.
Analysts undertaking that over 400 million AI robots will come on-line in simply the subsequent 4 years, and that the dimensions of the bodily AI business will eclipse €25 billion ($30 billion) over the identical time interval.
Not like LLMs, which have been skilled on the open web, bodily AI fashions should be taught from proprietary information, together with sensor feeds, video, robotic telemetry, edge circumstances captured within the subject and different sources. Storing and processing this information requires extra computational energy than storing and processing textual content.
Encord has seen demand surge as bodily AI strikes from experimentation to deployment:
- Knowledge on the corporate’s platform has grown from 1 petabyte to over 5 petabytes in twelve months – 3x greater than the info used to coach GPT-4
- Income from bodily AI clients has grown 10x over the identical interval
Eric Landau, co-founder and co-CEO of Encord, mentioned the funding will speed up product improvement and enlargement into new markets. “The businesses profitable in bodily AI perceive one thing that others are simply starting to appreciate: the mannequin is simply nearly as good as the info behind it. We’re constructing the infrastructure that makes that information usable – not simply as soon as, however constantly, as these methods be taught and enhance in the actual world.”
