Functional Architecture
Component | Description |
|---|---|
Connection | Communication link or interface between different software systems, allowing them to exchange data, perform actions or access resources. |
Dataset | Training or ingestion data from multiple sources like Storage Buckets, MYSQL, REST API. |
Pipelines | Training and inferencing pipelines using which models can be fine-tuned and trained and subsequently deployed and registered as an endpoint. |
Models | All models from the configured connections. The connections can be of type - OnPrem, AWS SageMaker, Azure ML, and GCP Vertex AI. |
Endpoints | All endpoints from the configured connections including REST APIs, model endpoints, etc. |
Adapters | Enable the integration and utilization of various services and capabilities of endpoints without providing host details. |
Remote Executor | Tool that allows the execution of pipelines or programs on a remote server or VM. |