Job Details

At Coco, our mission is to revolutionize urban logistics by empowering cities, boosting local economies, and delivering delightful customer experiences. We connect people with local merchants through our fleet of on-demand delivery robots, helping merchants reach their customers faster and more efficiently. By building innovative robotic systems that seamlessly navigate city sidewalks, Coco plays a key role in reshaping the future of last-mile delivery and enhancing local businesses.

To deliver on our mission, we are building an autonomy team to develop the AI technology that will enable our robot pilots to scale efficiently, sustainably, and safely. This involves building an autonomy stack ground-up based on our millions of miles of last-mile delivery routes, proprietary video streams, and LiDAR data.

What is the scope of this role?

As a Founding Machine Learning Engineer, you will be responsible for standing up Coco's autonomy stack alongside the CTO and fellow team members in the autonomy team. You will be responsible for designing, developing, and deploying our machine learning models as part of an end-to-end stack that feeds off millions of miles of daily piloted deliveries and sensor data (camera, LiDAR, etc.). The impact of this will be massive improvements to our robot-to-pilot ratio thereby allowing every person living in an urban area to benefit from last-mile delivery. In this role, you must accomplish the following:

• Develop the architecture and implementation of machine learning models that power the autonomous vehicle's perception, decision-making, and control systems.
• Design and implement algorithms for object detection, sensor fusion, path planning, and real-time decision-making.
• Develop and maintain robust data pipelines to ingest, label, and preprocess data from multiple sensors (LiDAR, cameras, radar).
• Train and optimize deep learning models using simulation and real-world data. Focus on improving system accuracy, robustness, and efficiency.
• Build simulation environments to test various driving scenarios and edge cases; work closely with the hardware team to ensure smooth integration of software with vehicle systems.
• Ensure that the autopilot system adheres to safety standards and regulations, incorporating redundancy and fail-safe mechanisms.
• Play a key role in recruiting, mentoring, and leading a high-performance machine learning and autonomy team.

• 7+ years experience in machine learning, computer vision, robotics, or autonomous systems.
• Proven track record of training and deploying a real world neural network.
• Extremely well versed in the models and techniques that have been used and are currently in use with autonomous driving or robotics systems in constrained, production environments. SfM, SLAM, large-scale mapping & localization, RL, imitation learning, neural rendering, etc.
• Expertise in training and deploying deep learning models, particularly in vision (CNNs, object detection, semantic segmentation).
• Excellent programming skills in Python and C++. Practical experience with PyTorch.
• Experience in optimizing large-scale ML models and deploying them on distributed cloud/edge infrastructure.
• Strong leadership and communication skills.

• Contributions to open-source autonomous vehicle frameworks (e.g., Autoware, Apollo).
• Experience with hardware integration and low-level system programming (ROS, CAN bus).
• Familiarity with reinforcement learning and imitation learning for decision-making in autonomous systems.
• Familiarity with regulatory and safety standards for autonomous vehicles (ISO 26262, UL 4600).