Software Engineer, ML Foundations Platform

As a Software Engineer at Zoox in the ML Foundations Platform role, you will design, develop, and deploy ML libraries and tools that enhance the development process. You'll collaborate with cross-functional teams to define requirements and align on architectural decisions, effectively reducing the time it takes to bring innovations from ideation to production.

To be considered for the Software Engineer position at Zoox, you should have at least 6 years of experience in software engineering, along with proficiency in Python or C++. Familiarity with ML frameworks and libraries such as PyTorch or TensorFlow is essential, as well as experience with GPU-accelerated inference on Nvidia hardware.

In your role as a Software Engineer on the ML Foundations Platform at Zoox, you will engage in projects that involve developing ML tools and libraries for autonomous driving, scene understanding, automated mapping, and model inference optimization. Each project will push the boundary of what's possible in autonomous vehicle technology.

At Zoox, Software Engineers in the ML Foundations Platform role have abundant opportunities for growth. As the company expands its robotaxi deployments, you'll be encouraged to explore new ML domains and technologies, with the support of a collaborative team environment focused on continuous learning and improvement.

The compensation for a Software Engineer position at Zoox includes a competitive salary range of $205,000 - $260,000, coupled with Amazon Restricted Stock Units and Zoox Stock Appreciation Rights. The package can also include a sign-on bonus, depending on factors like experience and location.

When discussing your experience with ML libraries and frameworks, be specific about the tools you're familiar with, such as PyTorch or TensorFlow. Highlight projects where you've used these frameworks, focusing on the challenges you faced and how you overcame them, especially in regards to model training and deployment.

In machine learning projects, debugging can be complex. Discuss your methodical approach, such as utilizing logging, visualization tools, or simplifying models to identify issues. Share examples where your debugging efforts enhanced model performance or resolved critical issues in the deployment process.

Explain how GPU acceleration significantly speeds up the training and inference processes in machine learning. Offer examples of frameworks that leverage GPU capabilities, such as TensorRT, and illustrate how using these tools impacted project outcomes by comparing it to CPU processing times.

Choose a project that showcases your ability to work in a team, emphasizing your role and contributions. Describe how you coordinated with other engineers or teams, the dynamics of collaboration, and how your team achieved project goals together while learning from each other.

Mention specific ways you keep up-to-date with the latest in machine learning, such as attending conferences, participating in online courses, or following relevant publications and thought leaders in the industry. Share how you apply new knowledge or techniques in your work.

Outline your approach by discussing the key components of an ML system for autonomous driving, such as data collection, model selection, testing, and deployment. Emphasize the importance of safety and reliability in your design process and how you would address these factors in real-world scenarios.

Discuss the tools you've used for version control, code reviews, and project management, such as Git, JIRA, or collaborative platforms like GitHub. Highlight how these tools have facilitated collaboration in your previous projects and improved team productivity.

Provide examples of challenges like data overfitting, model interpretability, or performance degradation in production. Discuss the strategies you employed to mitigate these issues, such as implementing regularization techniques, conducting extensive testing, and monitoring model performance over time.

Explain how the software development lifecycle is tailored for machine learning projects. Discuss the iterative nature of deploying ML models, including phases like data preparation, training, validation, and monitoring performance post-deployment, emphasizing how these differ from traditional software development.

Share your passion for machine learning and how it drives your desire to contribute to innovative projects, like those at Zoox that focus on autonomous driving. Discuss your excitement about shaping the future of mobility and how your values align with Zoox's mission.