Foundation models and autonomy software for industrial robots in unstructured environments
FieldAI builds the autonomy layer for robots operating in complex, unmapped environments—combining physics-based reasoning with learned models via their Field Foundation Model (FFM). The stack reveals dual engineering depth (PyTorch, TensorFlow, CUDA, Jetson for ML; ROS, C++, LiDAR for robotics) and production hardening (Kubernetes, Triton inference, edge optimization). Hiring velocity is accelerating across 47 open roles, with 22 engineering positions and a notable concentration of senior-level talent, signaling both technical complexity and a transition from research to field-deployed systems.
FieldAI develops autonomous robot software that works without pre-trained models, GPS, or pre-programmed routes. The company targets industrial operations—particularly construction—where robots encounter unpredictable, unstructured environments. Their core technology, the Field Foundation Model, learns from each deployment to improve fleet-wide autonomy and safety. The org is scaling rapidly: active projects span field deployment playbooks, enterprise customer success infrastructure, construction go-to-market motion, edge inference optimization, and 3D spatial data pipelines. Based in Irvine with 51–200 employees, they're hiring across the US, Japan, and Singapore.
PyTorch, TensorFlow, CUDA, Jetson for ML; ROS, C++, OpenCV, LiDAR, and Open3D for robotics perception; Triton and TensorRT for edge inference; RAG for retrieval-augmented generation.
Headquartered in Irvine, CA. Actively hiring in the United States, Japan, and Singapore across engineering, sales, support, marketing, and operations.
Other companies in the same industry, closest in size