Covariant

Covariant builds AI that enables robots to see, reason, and act in the real world, with a focus on warehouse and logistics automation. The company's foundation model for robotics, trained on massive datasets of real-world robotic interactions, allows robotic arms to pick, sort, and handle a vast diversity of items without being individually programmed for each object.

Unlike traditional industrial robots that follow rigid pre-programmed paths, Covariant's AI allows robots to handle novel objects, adapt to changing conditions, and learn from experience. This is critical in environments like e-commerce fulfillment centers where robots must handle millions of different products — from fragile electronics to flexible clothing.

Founded by AI researchers from UC Berkeley and OpenAI, Covariant has raised over $220 million and deployed its AI in warehouse operations for major logistics and retail companies. The company's approach of building a general-purpose robotic AI, rather than engineering solutions for specific picking tasks, positions it to scale across diverse industrial applications.