A prduction-ready ML pipeline using Tensorflow Extended (TFX), BigQuery, Apache Beam & Kubeflow on Google Cloud Platform. Components:
- Data generation with BigQuery ExampleGen
- Data streaming with Apache Beam
- Generate schema using schemagen
- Validator: Perform data validation
- Transformer: Transform into usable input for training
- Trainer: Train regression model
- Evaluator: Evaluate model performance metrics on eval set
- Pusher: Export model for serving
- Kubeflow backend for orchestrating pipeline