Kodiak builds autonomous trucking systems grounded in a heavy ML stack (PyTorch, TensorFlow, Ray, Kubeflow) paired with embedded systems tools (LIDAR, CAN, C++) for real-world vehicle control. The engineering-dominant org (43 of 73 roles) reflects active work on motion planning, safety case development, and multimodal transformer pipelines—core blockers in moving from simulation to fleet reliability. Hiring velocity is accelerating across senior and mid-level IC roles, signaling scaling from research into production autonomous operations.
Kodiak develops autonomous trucking solutions focused on Level 4 self-driving capability for commercial logistics. The company operates across three interconnected areas: the core autonomy system (perception, planning, control), fleet operations infrastructure (uptime, telemetry, continuous improvement), and training pipelines (leveraging real-world logs to scale model performance). Founded in 2018, Kodiak is a public company headquartered in Mountain View with 201–500 employees and hiring exclusively in the United States. Their technical challenges center on reducing latency in perception and planning, enhancing multi-target tracking, and building defensible safety cases for autonomous operation in complex driving scenarios.
Python, PyTorch, TensorFlow, C++, LIDAR, and embedded protocols (CAN, PTP). ML infrastructure includes Ray, Kubeflow, Dagster, and Apache Airflow. Visualization tools: Foxglove, Three.js, Plotly.
Mountain View, California. All hiring is currently in the United States.
Other companies in the same industry, closest in size