Thanks CloudBerry for your comment. The blog is just an example on how to use it with Spark. Primarily intended for existing Spark users with concurrent workloads. Certainly, you can achieve this using simple-salesforce and Kafka.