Streams

The platform distinguishes between two critical aspects of deploying machine learning models: handling incoming streaming data and managing model predictions. We refer to the processes that consume and manage streaming data as "Streams."

Management

The platform provides a dedicated interface for overseeing these system services, allowing users to efficiently manage and monitor the data flow into the platform.

Service

In the detail view, users can access a terminal session that displays real-time STDOUT from the process responsible for consuming streaming data and writing it to disk. This level of detail allows users to closely monitor the data ingestion process, observe the flow of data, and quickly identify issues or warnings that might arise.

By offering this transparent view into the streaming data process, LIT enables users to ensure that data is being captured and processed accurately, facilitating more effective management and troubleshooting of data streams. This comprehensive monitoring capability supports a smoother and more reliable data ingestion experience, which is crucial for maintaining the integrity and efficiency of real-time data workflows.