We are hiring a Program Manager to drive the expansion and integration of computer-vision-powered automation use cases across North America and embed them into our sustainment model. This role requires strong technical aptitude and data fluency. You will work hands-on with perception and detection metrics, sensor and image data, operational KPIs, and the ML pipeline lifecycle — from model training and evaluation through deployment and monitoring. You will partner daily with computer vision and ML scientists, engineers, RME teams, site operations, and external vendors to scope use cases, execute pilots, define and analyze KPIs, and drive data-informed decisions that scale reliably across a decentralized network. Key job responsibilities Identify and evaluate new computer vision automation use cases for North America (e.g., drones, robotics, autonomous vehicles, inspection systems); develop feasibility assessments and business cases for each opportunity. • Execute pilots end-to-end: define success metrics, coordinate cross-functional partners across science, engineering, RME, operations, and vendors, track milestones, and report outcomes with clear recommendations. • Conduct deep-dive analysis on model and detection performance (e.g., precision, recall, F1 score, detection accuracy); analyze flight telemetry, reliability logs, and operational data to identify trends and improvement opportunities. • Assess data and annotation pipeline quality feeding perception models; partner with science teams to evaluate labeling accuracy, dataset coverage, and annotation throughput. • Evaluate edge inference constraints including latency, on-device compute capacity, and connectivity limitations; factor deployment realities into use-case scoping and architecture decisions. • Support post-deployment model monitoring activities including tracking model drift indicators, retraining triggers, and A/B testing frameworks to ensure sustained system performance. • Build KPI dashboards and visualizations; translate telemetry and analytics into actionable insights for technical and non-technical stakeholders. • Coordinate with regulatory and safety stakeholders to ensure automation deployments comply with applicable standards (e.g., FAA regulations for UAV systems, robotics safety standards). • Translate between technical teams and operational stakeholders; document standards, RACIs, and handoff procedures so use cases remain sustainable post-launch. • Track risks, dependencies, and timelines across concurrent workstreams in a decentralized network environment. A day in the life • You Dive Deep into perception data and system telemetry to surface insights that others miss — then translate those insights into clear, actionable recommendations. • You demonstrate Bias for Action by driving pilots forward through ambiguity, coordinating across teams without waiting for perfect information. • You Earn Trust with cross-functional partners by communicating transparently, documenting decisions clearly, and delivering on commitments. • You Insist on the Highest Standards for data quality, pilot rigor, and documentation — raising the bar for how automation use cases are evaluated and sustained. LOCATION AND TRAVEL About the team Sustaining Technology Engineering (STE), within GES Sustainability & Automation, keeps Amazon’s deployed automation and robotics technologies reliable, standardized, and continuously improving across the network. Our portfolio spans drones, mobile robots, autonomous vehicles, and camera/sensor-based inspection systems — all powered by computer vision and machine learning. We partner with science, engineering, Reliability and Maintenance Engineering (RME), site operations, and vendor teams to take automation use cases from concept through sustained, scaled operation.