Building a Real-Time Analytics Stack with Apache Kafka and Apache Druid

Developing analytics applications for real-time monitoring and diagnosis of operational activity such as clickstreams or network flows is very different from building for traditional “BI” analytics. First, you need to execute analytics against a live data stream with minimal latency. Second, you need to store petabytes of real-time and historical data together. Third, you need to support hundreds to thousands of users. Last but not least, you need to support a variety of analytics methods, including real-time alerting, dashboards, ad hoc slice and dice, drill downs, search, ideally through a single UI. In this webinar, one of the creators of Apache Druid, will dive into: • The current state of real-time and streaming analytics, from stream processing like Spark and Flink to analytics tools including Tableau, Looker, Superset and Imply Pivot. • The challenges of architecting for real-time analytics, and how combining Kafka and Druid deliver low end-to-end latency at scale. • Design considerations and details behind Druid's integration with Kafka, Amazon Kinesis, and other messaging technologies for real-time ingestion at scale. • Architectural considerations for achieving speed and scale for real-time analytics including ad hoc, search and time-based queries.