Senior/Staff Machine Learning Engineer, Autonomy Validation

Zoox is on an ambitious journey to develop a full-stack autonomous mobility solution for cities and safely deploy a robotaxi service. Zoox’s System Design and Mission Assurance (SDMA) team is a foundational part of this journey, responsible for constructing the safety case and validating that our vehicles are safe for deployment. We are looking for a Senior or Staff Machine Learning Engineer to join our team to help improve our automated testing and validation processes.

This role is centered on applying cutting-edge machine learning to develop and enhance our validation processes. You will be instrumental in improving the efficiency and scalability of our testing by sampling across massive datasets where traditional methods no longer suffice. By working with both real-world fleet logs and synthetic data, your work will directly impact how we validate software changes and ensure our robotaxi service is both safe and reliable. Your work will also significantly contribute to the speed and efficiency of our validation process, allowing Zoox to go fast and achieve more.

You will be part of an organization with strong leadership and a transparent, respectful culture that enables you to reach your full potential. This high-impact position offers opportunities for career growth through demonstrated achievement.

• Lead Technical Initiatives: You will apply modern machine learning, including advanced techniques like encoder-decoder models, to critical validation problems at the intersection of ML and data science. You will serve as a key contributor and tech lead on a small, focused team.
• Improve Model Interpretability: You'll pioneer methods to understand the internal workings of our machine learning models, bridging the gap between "black box" models and systems engineering. This work will be crucial for building trust and ensuring the safety of our autonomous systems.
• Improve Feature Representation: You'll extend and refine the features and embedding space used by our models to better identify and cluster interesting driving scenarios. This involves leveraging a deep understanding of the autonomous driving stack and applying first-principles thinking to derive new, impactful features.
• Integrate AV Performance Data: You'll incorporate metrics and information on autonomous vehicle (AV) performance into the model to make its risk predictions more accurate and relevant.
• Optimize with Data Science: You'll apply your data science expertise to optimize models and sampling methodologies. This includes analyzing large datasets to identify patterns and critical insights that improve the model's ability to find risky or valuable scenarios for simulation.
• Collaborate Cross-Functionally: You'll work closely with system safety, data science, software, and fleet operations teams to understand their needs and integrate improvements that directly support our validation efforts.

• Experience: A PhD in a relevant field and/or 5+ years of experience working with machine learning models and data science methodologies in an industry setting.
• Technical Skills: Expertise in machine learning concepts, including model training, evaluation, and optimization. Strong programming skills in Python and experience with relevant machine learning libraries (e.g., PyTorch, TensorFlow, Jax). Experience with large-scale data processing and distributed computing.
• Domain Knowledge: Experience in robotics, autonomous vehicles, or a related field, with an understanding of challenges in perception, prediction, and planning.
• Mindset: Proven ability to drive progress independently, lead technical projects, and apply critical thinking to solve practical problems.
• Communication: Excellent communication skills and the ability to work effectively with cross-functional teams.

• Real-world impact as demonstrated in publications, patents, presentations, blog posts, etc. 
• Familiarity with encoder-decoder or foundation models for prediction and planning.
• Experience with test scripting and data analysis languages like SQL.
• Experience with techniques for machine learning model interpretability and explainability.
• Familiarity with the challenges of fleet data collection and validation in the autonomous vehicle space.