Machine Learning Automation Engineer

Job description

Within the System Behavior Analysis (SBA) organization, the Behavioral Automated Triage (BAT) team builds the ML, deep learning, and data pipelines that help Zoox understand and triage failures our robots encounter on the road. You will build agentic, LLM-powered triage systems and run them in production, keeping pipelines healthy that teams across Zoox use daily, while working closely with stakeholders across safety, planning, and quality assurance. In this role, you will: Design and improve ML/DL algorithms on large-scale data to automate test and triage workflows. Build and run agentic LLM systems (e.g., Claude or Gemini) that automate triage, from prototype to production. Run multiple production pipelines: monitor them, respond to issues, and fix the root causes. Work with stakeholders across data science, ML, autonomous drive planning, and quality assurance. Build monitoring tools and use them to keep improving your algorithms in production. Qualifications: PhD or Master's in a STEM field and 4+ years in Triage, Performance Analytics, for fortune 500 including infrastructure as code, ML, NLP, or deep learning. Strong productionization, CI/CD experience for medium to large projects Python,, with a solid grasp of dashboards, eval, algorithms and data manipulation. Experience with RAG, PyTorch, TensorFlow, or HuggingFace Has built or managed agentic LLM systems (Claude or Gemini) in a professional setting with large dataset to satisfy multiple stakeholders. Has run multiple production pipelines on platforms like Databricks and AWS, including fixing production issues. Bonus Qualifications: Experience with self correcting models, installation and maintenance of LLMs, data visualization tools like Looker/Dash/Databricks Dashboards Experience in Performance Monitoring at Scale, Support environments, autonomous driving, robotics. Strong verbal and written communication skills