Skip to main content

Introduction

This section is for workspace supervisors of the TANGO ecosystem, responsible for populating the TANGO platform with powerful AI capabilities. It explains what supervisors can do:

  • manage the users that can access to the supervisor workspace and assign them roles (USER or SUPERVISOR);
  • prepare, deploy and manage AI models within their designated workspaces.

The Model Lifecycle on TANGO​

Bringing a new model into the TANGO ecosystem involves managing its entire lifecycle. The workflow includes the following phases.

  1. Model Preparation: before a model can be deployed, it must be "packaged": its code, weights, dependencies and metadata must be structured in a standardized format that the TANGO platform can understand and execute;
  2. Deployment: alongside the model preparation, supervisors must create a deployment specifying where the model is stored and where it is served;
  3. Model registration: the model can then be registered in the TANGO platform, linking it to the deployment and making it available for use.

What You'll Find in This Section​

This part of the documentation provides the tools and guides for you to successfully build on TANGO.

  • Authentication and Users: learn how to authenticate with supervisor permissions and manage the users within your workspace, including inviting new members and editing roles.

  • Prepare a Model for TANGO: a guide that details the required structure and steps to package your custom model for compatibility with the platform.

  • Deployment Management: follow these instructions to tell TANGO where your model is stored and how it can be accessed. This includes setting up the model's serving endpoint and ensuring it is ready for use.

  • Storage Management: Learn how to configure the backend storage where data and file artifacts are kept.

  • Model Management: learn how to register your prepared model in TANGO, linking it to the deployment and making it available for users in your workspace.

  • OpenAPI and Postman Collection: access the tools to automate your build and management workflows via our comprehensive OpenAPI specification and a ready-to-use Postman collection.

Getting Started​

To begin, you must have been assigned a supervisor role for a workspace by a platform administrator, and you will need your personal login credentials.

Your journey starts with securing your access and setting up your team.