AI Geofencing Event Trigger Engineer

Location data becomes powerful when you can act on it in real time. A customer walking into a competitor's parking lot. A delivery truck detouring off the planned route. A field technician arriving at a service location. These events trigger business actions — a push notification, a routing update, a service log entry — but only if the geofencing system can detect them with sub-5-second latency at scale. We process location pings from 2.3 million opted-in devices per day and evaluate them against 180,000 active geofences for 240 enterprise clients. The current system handles the throughput. The failure modes are precision and recall: geofences that fire on nearby devices that haven't actually crossed the boundary, and boundary crossings that get missed during brief GPS dropouts. You'll improve the detection logic: building adaptive boundary tolerances based on device accuracy metadata, implementing trajectory prediction to fill in GPS gaps, and creating the monitoring that tells clients when their geofences are underperforming. The platform team is four engineers; you'll own the detection pipeline.