Introduction to Aidin
Aidin is a managed AI development and MLOps platform. It gives you JupyterLab notebooks, ML pipelines, a model registry, and managed inference endpoints on shared platform compute — without assembling or operating the infrastructure stack yourself.
What you get
Section titled “What you get”| Capability | Where it is documented | Availability |
|---|---|---|
| Interactive JupyterLab notebooks | Notebooks | Available |
| Workspace volumes and storage | Workspace storage | Available |
| ML pipelines (KFP) | Pipelines | Available |
| Model registry | Model registry | Available |
| Managed inference endpoints (KServe) | Model serving | Available |
| Hyperparameter search (Katib) | Integrations | Available |
| Accelerated GPU workspaces | Compute profiles | Limited — not offered self-service today |
| High-throughput LLM serving | Model serving | Coming soon |
| Serverless scale-to-zero endpoints | Model serving | Coming soon |
How these docs are written
Section titled “How these docs are written”Aidin docs publish verified platform behavior only. Every page marks its verification state:
- Verified statements were checked against the deployed platform and are safe to rely on.
- Pending verification sections describe planned or in-progress documentation. Behavior stated there is not yet confirmed and may change.
Where Aidin’s behavior differs from the upstream open-source projects it builds on (Kubeflow, KServe, Katib), Aidin’s documented behavior takes precedence.
Getting around
Section titled “Getting around”- Create your account — sign up and verify your email.
- Your workspace — what Aidin provisions for you automatically.
- Your first notebook — from sign-in to a running JupyterLab session.
- Your first model deployment — the path from notebook to endpoint.
For compute, quota, and scheduling behavior see Cloud resources. For plan and usage questions see Billing.