Join Amazon's Last Mile Geospatial Technology team and help build the intelligence layer that powers millions of deliveries every day. The Geospatial organization within Last Mile Technology owns the foundational location and mapping systems that Amazon Logistics depends on — from geocoding and address resolution to delivery accuracy and spatial analytics. The team is also responsible for closing the loop between delivery outcomes and geospatial data quality, driving continuous improvement across the delivery network. We're building an agentic system that automatically investigates delivery issues, correlates signals across geocode data, address metadata, delivery patterns and history, and produces actionable insights at scale. Think of it as an always-on geospatial analyst that can trace a wrong-to-address delivery back through layers of spatial data, identify root causes, and recommend fixes — all without human intervention. This is a unique opportunity to work at the intersection of geospatial systems, data engineering, and agentic AI. You won't just be writing CRUD APIs — you'll be designing autonomous investigation workflows, building correlation engines across heterogeneous spatial datasets, and creating systems that reason about real-world delivery problems. Key job responsibilities - Design and build core components of the agent — investigation pipelines, signal correlation logic, and automated root-cause analysis workflows - Work with geospatial data at scale: geocodes, address graphs, delivery traces, proof-of-delivery signals, and map layers - Build integrations with internal geospatial platforms to power automated lookups and analysis - Develop data pipelines that ingest, transform, and connect delivery outcome data with spatial quality metrics - Collaborate with applied scientists and geospatial engineers to translate investigation heuristics into scalable, automated systems - Own your services end-to-end — from design through deployment, monitoring, and operational excellence - Define system architectures and designs, exploring technical feasibility tradeoffs, and making data-driven decisions. - Help drive the architecture and technology choices that enable a world-class experience A day in the life As an SDE on GeoLoop, you'll build Chronicles — an agentic system that automatically investigates delivery issues by reasoning across Amazon's geospatial data. Day-to-day, you'll design autonomous investigation workflows, write correlation logic connecting geocode quality and delivery outcomes, and ship features that directly reduce wrong deliveries. You'll collaborate with applied scientists, geospatial engineers, and operations teams who rely on your system to surface root causes at scale. The end customer is the Amazon delivery customer — every automated investigation improves the spatial data that powers their experience. Small team, end-to-end ownership, real architectural influence.