Low-latency Inference Using Databricks ML in StreamSets

In my previous blog, we looked at using TensorFlow models in dataflow pipelines to generate predictions and classifications in real-time. In this blog post, I will walk you through using Databricks ML models in StreamSets Data Collector for low-latency inference. In Machine Learning there are two major phases. This can be broken down in two categories. Like TensorFlow Serving, MLeap, and PMML, Databricks ML Model Export is also targeted at low-latency, lightweight ML-powered applications.