A geofence is a virtual perimeter for a real-world geographic area. Geofencing is a technique used to monitor a geographical area by dividing it into smaller subareas demarcated by geofences. It can be used to create triggers whenever a device moves across a geofence to provide useful location-based services. Since real-world objects tend to move continuously, it is essential to provide these services in real-time to be effective. Towards this objective, this paper introduces a scalable data pipeline for geofencing that can reliably handle and process data streams with high velocity using Apache Pulsar - an open-source Publish/Subscribe messaging system that has both stream processing and light-weight computational capabilities. Further, an implementation of the proposed data pipeline for a specific real-world case study is presented to demonstrate the envisaged advantages of the same.