Machine Learning Streaming

A Machine Learning streaming pipeline with a user interface developed in an open source environment such as Kafka (KRaft mode), Debezium, Postgres and BentoML. This project is living in AWS EC2 instances, using Terraform, Ansible and Docker, so its deployment and replication capacity is straighforward.

Features

• Full-stack app in NodeJS and React with a NGINX server
• Kafka platform in KRaft Mode
• Kafka Connect and Avro Schema Registry
• Terraform, Ansible and Docker
• PostgreSQL
• Debezium and JDBC connectors
• Scripts made in Bash, YAML and Python
• ML classification model use case (a churn example)
• BentoML
• Cloud computing in AWS, IaaS approach

Authors

• @roniepaolo

References

• Machine Learning Streaming with Kafka, Debezium and BentoML