User stories

Jan 5, 2021

Lyft’s modern data architecture feat. Apache Druid, Kafka and Flink

Take a look at Lyft’s modern data architecture. They have an app implemented in Flink that reads real-time event data from Kinesis and transforms the data. It dumps the event data out in protobuf onto a Kafka topic and Druid consumes the events from Kafka. From there, Druid begins reading the data in real-time, which will then be available for a user to start querying via Superset.