KSQL UDF for sensor analytics. Leverages the new API features of KSQL to build UDF / UDAF functions easily with Java to do continuous stream processing with Apache Kafka. Use Case: Connected Cars – Real Time Streaming Analytics using Deep Learning.
Machine Learning / Deep Learning models can be used in different ways to do predictions. Natively in the application or hosted in a remote model server. Then you combine stream processing with RPC / Request-Response paradigm. This blog post shows examples of stream processing vs. RPC model serving using Java, Apache Kafka, Kafka Streams, gRPC and TensorFlow Serving.
I am happy that my first official Confluent blog post was published and want to link to it from by blog:
How to Build and Deploy Scalable Machine Learning in Production with Apache Kafka
I do a lot of presentations these days at meetups and conferences about how to leverage Apache Kafka and Kafka Streams to apply analytic models (built with H2O, TensorFlow, DeepLearning4J and other frameworks) to scalable, mission-critical environments. As many attendees have asked me, I created a video recording about this talk (focusing on live demos).
After three great years at TIBCO Software, I move back to open source and join Confluent, the company behind the open source project Apache Kafka to build mission-critical, scalable infrastructures for messaging, integration and stream processsing. In this blog post, I want to share why I see the future for middleware and big data analytics in open source technologies, why I really like Confluent, what I will focus on in the next months, and why I am so excited about this next step in my career.