Skip to content

User Interface

Workspaces Landing Page

Upon entering the platform, users are greeted with a landing page displaying a table of all existing workspaces. For the creation of new workspaces, a conveniently placed "Create" button is available.

Workspace overview

Selecting any workspace takes users to the workspace page.

Workspace Page

Experiment tracking is by default the first page within the workspace. This pages shows existing experiments in the workspace.

The workspace page has a left sidebar, serving as the navigation hub for various tools and features.

Left bare

  • Tracking: The tracking system is a robust suite for managing machine learning experiments and model runs. It offers detailed logging and comparison tools, essential for experiment analysis.

  • Models: This section simplifies model registration and configuration. Users can easily prepare models for deployment, streamlining the transition from development to production.

  • Deployments: Deployment tools to facilitate the launching and monitoring of models in production environments, ensuring smooth and efficient model rollouts.

  • Jobs: This moodule allows using the platform with custom training and configuration files. This helps developers train, monitor and utilize the resources within the platform as well as being flexible with their model training tasks.

  • Workflows: This module is a custom API-triggered backend workflows builder. It facilitates the request of various API requests.

  • LLM: This dedicated module facilitates the LLM Fine-tuning and benchmarking using pre-built tools, as well as connecting various benchmarking tools and datasets.

  • Annotations: This section helps feed the training data with the human feedback and AI agents to annotate the data.

  • Blueprints: This module helps users create blueprints such that they can repeat the custom workflows, secrets and settings and utilize in their next steps.

  • Datasets: A comprehensive section dedicated to the creation, management, versioning, and organization of machine learning datasets. It supports maintaining data integrity and consistency across projects.

  • Notebooks: Integrated directly within the platform, the Notebook page features a Jupyter notebook environment. It offers a seamless experience for conducting analyses and running code within the workspace.

  • Usage Monitoring: This is a dedicated admin tool that helps tenant monitor the usage of resources across workspaces, users, workloads.

  • User Governance: This feature provides efficient access management, allowing for precise control over workspace ownership and permissions.

  • Settings: This section is dedicated to the management of workspace settings, enabling users to customize their environment according to their specific needs.