Utilizing the FLiP stack we can drive NLP and other machine learning classifications in the cloud, on-premise, hybrid and at the edge utilizing the open source Apache Pulsar project. We often utilize the full FLiPN stack which includes Apache Flink, Apache NiFi and Apache Pulsar.
Apache Pulsar is a unified messaging platform that can host and execute classifications as events arrive asynchronously to topics. We deploy our functionality in lightweight functions that can run as processes, threads or in K8. We often use the serverless FunctionMesh to run these. This allows us to run machine learning classes built in Go, Python and Java. We will show you utilizing Vader Sentiment, Pytorch, TensorFlow, DJL.AI, MXNet and other machine learning options with ease.