Aizen Platform Interfaces
This diagram shows the interfaces to the Aizen platform.
There are three interfaces into the Aizen platform.
Data Sources
This interface connects your database and streaming data sources to the Aizen platform. Databases are connected as JDBC endpoints. Stream sources are connected as Kafka endpoints. There are methods to connect CSV files, Snowflake, and Cloud data sources as well.
Prediction Request Sources
This interface connects your prediction application to the Aizen platform. Prediction applications make REST requests to an http endpoint in the Aizen platform. There are mechanisms to connect batch sources such as databases and Cloud data sources for batch predictions, as well as Kafka endpoints for predictions on streaming data.
Job Management
You manage machine learning jobs via a Jupyter notebook connection to the Aizen platform. You may run jobs to create ML features from your data, train an ML model or deploy a trained ML model for prediction.
Last updated