Skild AI
Skild AI is building a universal robotics foundation model that controls arms, humanoids and quadrupeds through simulation-trained, cross-embodiment machine learning.
Skild AI, founded in 2023 by Carnegie Mellon professors Deepak Pathak and Abhinav Gupta, is building a foundation model for robotics: a universal robot brain designed to control robotic arms, humanoids and quadrupeds alike. The model learns from observation and adapts to new bodies and environments, drawing on massive simulation, machine learning and techniques for sim-to-real transfer and cross-embodiment control.
The technical work spans deep learning, computer vision and physically-grounded intelligence. Rather than training separate controllers for each machine, the company's approach targets a single model capable of generalising across embodiments - a problem set that includes learning from observation and closing the gap between simulation and the physical world. Applications of interest include manufacturing, logistics and industrial automation.
Skild AI raised $300m in a Series A round and counts NVIDIA and Amazon among its named partners. The company describes its ambition as building for robots what OpenAI has built for digital intelligence. It presents itself as a fast-moving, mission-driven team that values technical excellence and collaborative problem-solving, and it recruits engineers and researchers who want to see their algorithms run in the physical world rather than only on servers.