Applied Science Manager, Prime Air

Amazon
Seattle, Washington, USA2026-04-20ONSITE

About the job

As an Applied Science Manager, you will lead a high-caliber team of scientists and engineers focused on the 'brains' of our fleet orchestration. You will bridge the gap between Geospatial Intelligence and Machine Learning to revolutionize our path planning and scheduling algorithms. Your primary north star? Increasing deliveries per hour (DPH) through intelligent, automated optimization.

Responsibilities

Lead & Mentor: Manage a cross-functional team of Applied Scientists and Engineers, fostering a culture of scientific rigor and rapid iteration.

Innovate Path Planning: Leverage ML/RL and heuristic search techniques to develop dynamic path-planning algorithms that navigate complex airspace and weather patterns.

Optimize Orchestration: Drive the development of high-scale scheduling systems that manage battery life, maintenance cycles, and delivery windows to maximize fleet utilization.

Geospatial Mastery: Utilize deep geospatial data (3D maps, urban topology, etc.) to improve situational awareness and mission safety.

System Architecture: Define the long-term technical roadmap for mission orchestration, ensuring our systems are modular, scalable, and resilient.

Cross-Functional Collaboration: Partner with Hardware, Flight Safety, and Supply Chain teams to translate business requirements into technical breakthroughs.

Qualifications

Minimum

Experience managing a team of scientists and/or engineers in a production environment.

PhD or Master’s degree in Computer Science, Robotics, Operations Research, or a related field.

Strong foundation in Geospatial Information Systems (GIS) and spatial data analysis.

Proven track record of applying Machine Learning (e.g., Reinforcement Learning, Graph Neural Networks) to optimization problems like path planning or vehicle routing.

Preferred

Experience with autonomous systems, UAVs, or large-scale logistics networks.

Knowledge of combinatorial optimization and real-time scheduling constraints.

A knack for turning ambiguous 'blue sky' research into deployed, high-impact features.