Build production AI at the intersection of LLMs, edge ML, and sensor fusion — Navigation Intelligence that directly impacts the safety and experience of millions of Amazon transporters every day.

We're seeking a Software Development Manager to lead the Navigation Intelligence team in Austin. The team owns the navigation experience in Amazon's delivery associate app, building a hybrid on-device/cloud AI architecture that combines low-latency on-device inference with fleet-wide pattern analysis and LLM-powered contextual guidance. This is production AI at scale: billions of requests monthly, safety-critical latency requirements, and direct impact on transporter safety and experience.

The SDM will lead a team of software engineers building systems that analyze transporter behavior patterns to anticipate navigation issues before they occur, deliver clear and natural guidance through LLM-optimized directions, and provide context-aware error handling that replaces cryptic messages with actionable insights. The role requires balancing on-device constraints (battery, compute, memory) with cloud capabilities (scalability, ML pipelines, LLM services) while ensuring every feature ships with transporter safety as the top priority.

Key responsibilities include:
- Leading a team of software engineers building navigation systems used by millions of transporters globally
- Driving technical strategy for hybrid on-device/cloud AI architecture with low latency requirements
- Partnering with routing optimization teams, platform/operations teams, and Science teams to deliver integrated navigation experiences
- Working with route planning teams and Safety teams who depend on navigation accuracy metrics
- Owning problem domains including pre-trip intelligence, real-time traffic detection and rerouting, location accuracy through sensor fusion, arrival/parking guidance, and alert prioritization
- Managing technical roadmaps that balance innovation with operational stability
- Developing engineering talent and building high-performing teams
- Writing narratives and influencing VP-level technical decisions

The technical environment includes on-device development (iOS/Android, sensor fusion, edge ML models), cloud services (AWS, distributed systems, ML pipelines, LLM integration), and real-time streaming architectures. Scale includes billions of navigation requests monthly, 22.5B+ daily route calculations from upstream routing services, and global deployment across diverse geographies and network conditions.

Amazon's Geospatial teams operate with a safety-first culture. The environment requires someone who can ship production AI systems at scale, is comfortable navigating ambiguity and cross-team dependencies, and is passionate about applying emerging AI capabilities to hard, real-world problems. 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 building navigation systems used by millions of transporters globally
- Driving technical strategy for hybrid on-device/cloud AI architecture with low latency requirements
- Partnering with routing optimization teams, platform/operations teams, and Science teams to deliver integrated navigation experiences
- Working with route planning teams and Safety teams who depend on navigation accuracy metrics
- Owning problem domains including pre-trip intelligence, real-time traffic detection and rerouting, location accuracy through sensor fusion, arrival/parking guidance, and alert prioritization
- Managing technical roadmaps that balance innovation with operational stability
- Developing engineering talent and building high-performing teams
- Writing narratives and influencing VP-level technical decisions

About the team
The Navigation Intelligence team owns the navigation experience in Amazon's delivery app, used by millions of transporters globally. We build the systems that guide transporters from stop to stop — turn-by-turn directions, hazard detection, rerouting, parking guidance, and contextual alerts that reduce cognitive load during delivery. Our metrics directly impact transporter safety and delivery performance.