Software Development Manager, Geospatial

Build the routing engine behind every Amazon delivery. Planet-scale pathfinding that runs a trillion route calculations a day at microsecond latency, shaping how millions of packages reach customers safely and on time, every day. We're seeking a Software Development Manager to lead the Transit Intelligence team in Bellevue. The team owns the road network and routing foundation for Amazon Last Mile: road network construction and vending, pathfinding and transit time estimation. These systems compute how long it takes to get from A to B and what the safest path is, at two extremes of scale at once. Navigation needs answers in milliseconds in real time; route planning needs the full origin-by-destination product, adding up to over 1.2 trillion pathfinding calculations across the fleet each day. The SDM will lead a team of software engineers building the algorithms and infrastructure behind this scale: effective precomputation, succinct graph representations and the caching, sharding, and vending strategies that turn a continent-scale road graph into microsecond lookups. SDM will be responsible for logic in service that fuses real-time traffic with historical movement profiles and returns time, distance, and a path risk signal, so the routes it produces optimize for safety alongside speed. The role requires someone who can hold both the low-latency navigation workload and the high-throughput planning workload in view at once. Key responsibilities include: - Leading a team of software engineers who own road network construction and vending, pathfinding, and transit time estimation for Amazon Last Mile. - Partnering with Planning systems teams who solve the Traveling Salesman Problem at very high throughput, and owning the API contracts and accuracy bar those solvers depend on. - Working with Safety teams on routing and path-risk accuracy. - Partnering with Science teams on the models behind time estimation and integrating those models into production routing. - Owning transit time estimation quality end to end, including planned-versus-actual analysis and the underestimation and overestimation that drive downstream planning and customer promise. - Managing technical roadmaps that balance new capability against the operational excellence a Tier-1 service demands. - Developing engineering talent and building a deep, resilient team with clear technical ownership. - Writing narratives and influencing VP-level technical decisions. The technical environment includes large-scale distributed systems, graph algorithms and pre-computation pipelines, real-time data ingestion, AWS, and ML model integration. Scale includes continent-scale road networks, Tier-1 availability and latency SLAs. Amazon's Geospatial teams operate with a safety-first culture. The role requires someone who can operate a high-throughput, low-latency service at scale and is comfortable navigating ambiguity and dense cross-team dependencies. This role reports to a Senior Software Development Manager within the Geospatial organization. Key job responsibilities Key responsibilities include: - Leading a team of software engineers who own road network construction and vending, pathfinding, and transit time estimation for Amazon Last Mile. - Driving technical strategy for precomputation, graph representation, and the caching, sharding, and vending architecture behind both the low-latency navigation workload and the high-throughput planning workload. - Partnering with Planning systems teams who solve the Traveling Salesman Problem at very high throughput, and owning the API contracts and accuracy bar those solvers depend on. - Working with Safety teams on routing and path-risk accuracy. - Partnering with Science teams on the models behind time estimation and integrating those models into production routing. - Owning transit time estimation quality end to end, including planned-versus-actual analysis and the underestimation and overestimation that drive downstream planning and