Senior Machine Learning Engineer, Vehicle Behaviors

About Stack: Stack is developing revolutionary AI and advanced autonomous systems designed to enhance safety, reliability, and efficiency of modern operations. Stack's autonomous technology incorporates cutting-edge advancements in artificial intelligence, robotics, machine learning, and cloud technologies, empowering us to create innovative solutions that address the needs and challenges of the dynamic trucking transportation industry. With decades of experience creating and deploying real world systems for demanding environments, the Stack team is dedicated to developing an autonomous solution ecosystem tailored to the trucking industry's unique demands. About the Team The Vehicle Behaviors team is responsible for improving the decision-making capabilities and reliability of the AV in real world scenarios. Working broadly across domains and planning components, this team owns everything from ML-based Agent Predictions and Ego Trajectory Generation to safety-critical and heuristic-driven components like Trajectory Selection, to off-board metrics and tools used to measure performance and prevent regressions. This team has some of the most visible, high-impact work on the AV’s performance and plenty of challenging problems still to solve. About the Role We are looking for strong machine learning engineers to help in developing and deploying motion planning components for next generation self-driving systems. This requires strong ownership over critical components that cross domain boundaries. Candidates need to understand, experiment, improve, and field state-of-the-art machine learning systems in real-time, safety-critical applications. We are focused on building a product. Candidates should have a mission-driven mindset and customer-centric obsession to deliver a compelling product, and be able to work with significant cross-functional interactions. You might be a good fit for this role if you have experience with any of the following: Deploying state of the art ML models or architectures to solve planning problems on fielded products Refining ML outputs for use in down-stream planning components Integrating ML-based planning techniques side-by-side with heuristic approaches in production Building data mining, data introspection, or other critical ML tooling What Success Looks Like Responsibilities: Design, scope, implement, and integrate machine learning systems to solve on-vehicle behavior problems in a real-time, resource-constrained environment. Develop tooling and infrastructure necessary to measure and iterate on ML models. Work closely with systems engineers to ensure a safe, well tested product is delivered. Work closely with verification teams to ensure proper testing and validation of the motion planning modules. Make extensive use of unit testing, simulation, and log simulation to properly validate their work. Collaborate with other autonomy teams including perception, localization, controls, etc. to ensure solutions are appropriate for real-world performance Identify bottlenecks and limitations in system performance, and develop novel motion planning components to unlock new capabilities and ensure a reliable system. Be involved in experimentation, design and iteration exercises, and help to align stakeholders by using strong presentation and communication skills. Experience: Strong Python skills Experience with key ML libraries like PyTorch and TensorRT Experience with cloud orchestration tools like Flyte Strong written and verbal communications skills #LI-AW1 We are proud to be an equal opportunity workplace. We believe that diverse teams produce the best ideas and outcomes. We are committed to building a culture of inclusion, entrepreneurship, and innovation across gender, race, age, sexual orientation, religion, disability, and identity. Check out our Privacy Policy. Please Note: Pursuant to its business activities and use of technology, Stack AV complies with all applicable U.S. n ... (truncated, view full listing at source)