Senior AI/ML Engineer

Senior Ai/Ml Engineer

The incumbent supports the modernization of CPSC's analytical ecosystem by designing and implementing AI/ML models, advanced analytical methods, cloud native workflows, and automation that accelerate hazard triage and strengthen data driven decision making across the agency.

The Senior AI/ML Engineer performs the following duties:

• Designs, develops, tests, and deploys supervised, unsupervised, NLP, deep learning, and agent-enabled AI models that detect product hazards, surface signals, and identify leading indicators.
• Uses modern ML frameworks (e.g., PyTorch, TensorFlow) and scalable compute environments to build high-performance models that meet standards for accuracy, fairness, robustness, and transparency.
• Designs and implements high-quality data pipelines for ingestion, transformation, validation, quality enforcement, and preprocessing of diverse datasets.
• Operationalizes models in secure cloud environments (Azure), leveraging cloud-native services, containerization, CI/CD automation, and scalable compute.
• Conducts analytical research to identify trends, patterns, and emerging hazards through statistical learning, causal inference, scenario modeling, and "what-if" simulations.
• Supports the agency's hazard detection research agenda by proposing new analytical methods, experimenting with novel architectures, and implementing advanced machine learning approaches.
• Develops and maintains state-of-the-art knowledge in AI/ML, participating in working groups, communities of practice, and interagency collaborations.
• Works closely with program offices, engineering teams, compliance staff, field investigators, and agency leadership to translate analytic requirements into operational ML solutions.
• Communicates uncertainty, assumptions, limitations, and risk tradeoffs associated with modeling decisions and analytical findings.
• Designs and develops AI agents and retrieval-augmented systems (e.g., Copilot Studio, Python-based agents) that operationalize hazard detection workflows.
• Integrates AI agents with APIs, event streams, dashboards, and case management systems to reduce cycle time from signal to action.
• Performs other related duties as assigned.