Jan Kesters

In this session we will explain how to deploy a Deep Learning model and use it to predict real-time the location of a device. Kafka will be used to ingest data from a device and to collect real-time distance data from cellphone towers. We’ll explore two possible ways to use Kafka services in Azure either via Docker containers or leveraging the Kafka endpoints within Azure Event hubs. We will show-case how to use Scala for building an event processing pipeline that prepares data for both training and prediction by the model, and how to deploy the model in production for real-time inference. Key technologies used are: Azure Docker Hub, Azure Event Hubs, Azure SQL DB, Jupyter Notebooks using Python.