The following diagram shows the high level architecture of Odahu-flow project.
- Odahu-flow CLI is a command-line interface for interacting with Odahu-flow API service.
- The Odahu-flow Swagger UI provides swagger interface for Odahu-flow REST API.
- Odahu-flow Airflow plugin provides a set of custom operators that allow you to interact with a Odahu cluster using Apache Airflow
- The MLflow Tracking component is an API and UI for logging parameters, code versions, and metrics when running your machine learning code and for later visualizing the results.
- JupyterLab extension allows you to interact with an Odahu cluster from JupyterLab web-based IDEs.
- API service manages Odahu Platform entities: Connections, Trainings, Packaging, Deployments.
- Service catalog provides a Swagger UI for Model Deployments.
- TektonCD is an open source implementation to configure and run CI/CD style pipelines for your Kubernetes application.
- Knative Serving builds on Kubernetes and Istio to support deploying and serving of serverless applications and functions.
- Odahu-flow Model Training API provides features to manage remote training jobs.
- Odahu-flow packagers turn a Trained Model Binary artifact into a specific application.
- Odahu-flow Model Deployment API allows deploy ML models in a Kubernetes cluster.