<p><strong><span class="legendSpanClass">SAN FRANCISCO</span></strong> &#8212; Encord, a data infrastructure company for physical AI, has announced a $60 million Series C led by Wellington Management, bringing the company&#8217;s total funding to $110 million. Existing investors Y Combinator, CRV, N47, Crane Venture Partners and Harpoon Ventures also participated in the round alongside new investors Bright Pixel Capital and Isomer Capital.</p>
<p>The investment will help Encord scale its AI-native data infrastructure platform, which helps AI teams manage, curate, annotate, and align the multimodal data that physical AI systems depend on, including audio, video, images, sensor data, 3D point clouds and other formats that legacy data platforms weren&#8217;t built to handle.</p>
<div class="pull-right inline-gallery-container col-md-8 col-sm-7 col-xs-12">
<div class="row">
<div class="col-sm-12">
<aside class="pull-quote">Encord&#8217;s AI-native data infrastructure manages, curates, annotates, and aligns multimodal data for physical AI systems.</p>
</aside>
</div>
</div>
</div>
<p>Encord works with over 300 AI teams globally, including Woven by Toyota, Skydio, AXA Financial and numerous physical AI and frontier labs. The company has seen significant growth in both revenue and data volume on its platform in the last twelve months as a result of the surge in physical AI.</p>
<p>Encord&#8217;s Series C comes as physical AI &#8211; which powers robots, autonomous vehicles, drones, and other systems that operate in the real world &#8211; enters an explosive new growth stage. After years of lab demos and pilot programs, these systems are moving into production. Analysts project that over 400 million AI robots will come online in just the next 4 years, and that the size of the physical AI industry will eclipse $30B over the same time period.</p>
<p>Unlike large language models, which were trained on the open internet, physical AI models must learn from proprietary data, including sensor feeds, video, robotic telemetry, edge cases captured in the field and other sources. Storing and processing this data requires more computational power than storing and processing text.</p>
<p>That data doesn&#8217;t organize itself. Getting the right data into the models and keeping the wrong data out—continuously, at scale—requires purpose-built AI-native data infrastructure.</p>
<p>&#8220;Everyone is focused on building bigger models,&#8221; said Ulrik Stig Hansen, Co-Founder and Co-CEO of Encord. &#8220;But for physical AI, the bottleneck isn&#8217;t model size. It&#8217;s data readiness. You can have the most sophisticated model in the world, and it will still fail if the data feeding it is incomplete, inconsistent, or misaligned with real-world conditions. That&#8217;s the problem we solve.&#8221;</p>

NVIDIA, at its GTC conference, has announced the NVIDIA NemoClaw stack for the OpenClaw agent platform…
Super Micro Computer shares plunged by 33% Friday after a co-founder of the company was…
SAN JOSE -- PayPal is making PayPal USD (PYUSD) available in 70 markets worldwide in…
Uber Eats has raised commission fees for restaurant orders on its marketplace. The changes affect…
PALO ALTO -- Genspark.ai announced the launch of Genspark Claw, introduced as users’ first “AI employee.”…
SAN FRANCISCO -- Quince, a consumer technology platform redefining how premium goods are produced, priced,…