European robotics startup Microagi is partnering with Google Cloud to accelerate the development of AI models that allow robots to better understand and interact with the physical world.
The collaboration gives Munich-based Microagi access to Google Cloud’s AI infrastructure and NVIDIA Blackwell-powered computing resources to train what the company calls “embodied AI” models, AI systems designed specifically for robots operating in real-world environments.
The partnership is intended to help Microagi expand its enterprise robotics platform while speeding the deployment of AI-powered robots across industries including manufacturing, hospitality and logistics.
Google Cloud Powers Large-Scale AI Training
As part of the agreement, Microagi will use Google Cloud’s advanced AI platform alongside NVIDIA Blackwell GPUs and rack-scale systems to process massive multimodal datasets that include video and other physical-world information.
The infrastructure will support both AI model training and inference, enabling Microagi to develop task-specific software packages for different types of commercial robots.
The company also plans to use Google’s Gemini Enterprise Agent Platform as part of its AI development workflow.
Building AI for Real-World Robots
Unlike general-purpose AI models, Microagi focuses on creating customized AI systems tailored to specific robotic platforms and operational tasks.
Its Atlas platform fine-tunes AI models using a customer’s own operational data, allowing businesses to deploy robots configured for roles such as warehouse operations, manufacturing processes or hospitality services.
Because the platform is hardware-agnostic, organizations can integrate multiple robotics systems without becoming dependent on a single AI vendor.
Enterprise Robotics Continues to Grow
Demand for intelligent robots is increasing as organizations look to automate repetitive tasks while addressing labor shortages and improving operational efficiency.
Training AI models for robotics requires significantly more computing power than traditional AI applications because robots must interpret video, movement, spatial awareness and real-time environmental changes simultaneously.
The collaboration between Microagi, Google Cloud and NVIDIA is designed to provide the computing infrastructure necessary to support those increasingly complex workloads.
The Bottom Line
Enterprise robotics is moving beyond research labs and pilot projects into real-world commercial deployments. By combining Google Cloud’s AI platform, NVIDIA’s accelerated computing technology and Microagi’s robotics software, the companies aim to make intelligent robots more practical and scalable for businesses across multiple industries.