Tetra Data & AI Workspace

🚧

IMPORTANT

Tetra Data & AI Workspace is available through an early adopter program (EAP) currently and is activated for customers through coordination with TetraScience. The feature may require changes in future Tetra Data Platform (TDP) releases. For more information, or to activate the Tetra Data & AI Workspace in your TDP environment, contact your customer success manager (CSM) or account executive.

Tetra Data & AI Workspace helps you colocate your replatformed and engineered scientific data with analysis applications, and then correlate analysis results through Tetra Data Apps—all while maintaining a source of record in the Tetra Scientific Data and AI Cloud.

By using the Data & AI Workspace page to access your organization's configured Tetra Data Apps, you can do the following:

  • Visualize and analyze scientific data to generate analysis results
  • Verify the data as part of the automated scientific workflow
  • Easily associate primary data with analysis output
  • Annotate scientific data to support AI/ML model training
  • Generate predictions and performance AI/ML model inference and continuous learning

Working With Tetra Data Apps

There are two main types of Tetra Data Apps:

  • Linked Data Apps run as external SaaS applications outside of the Tetra Data Platform (TDP).
  • Embedded Data Apps run in the cloud and are configured for SaaS application streaming back to the TDP. Some Embedded Data Apps are based on Streamlit, while others are based on Windows. The user experience for each Embedded Data App type is slightly different, but the process for managing them is the same.

Your organization's analysis software determines the Data App type that you use. To configure Tetra Data Apps for an organization, see Tetra Data Apps.

📘

NOTE

The Tetra Data & AI Workspace and Tetra Data Apps don't use artificial intelligence (AI) or machine learning (ML) natively. Instead, both features support AI/ML model training in third-party applications that run outside of the TDP. For Embedded Data Apps based on Streamlit, TetraScience has the ability to create Python code and make it available to customers through the TDP. In the future, code may be created to help integrate customers' Tetra Data with AI/ML models on third-party applications. However, customers can't modify this code, and any code developed in this way must first go through our ISO 9001-certified Quality Management System and software development lifecycle.

Access the Data & AI Workspace Page

📘

NOTE

If Data & AI Workspace doesn't appear in the left navigation menu, the feature hasn't been activated for your organization. To activate the Tetra Data & AI Workspace for your organization, please contact your CSM.

To access the Data & AI Workspace page, do the following:

  1. Sign in to the Tetra Data Platform (TDP).
  2. In the left navigation menu, choose Data & AI Workspace. The Data & AI Workspace page appears and displays two tabs:
    • Apps shows your organization's configured Tetra Data Apps.
    • Sessions shows your organization's previous Data & AI Workspace sessions.

Tetra Data & AI Workspace

Start a New Data & AI Workspace Session

You can start a new Data & AI Workspace session from either the Data & AI Workspace page or, for Embedded Data Apps based on Windows, by selecting a set of files from the Search page.

Start a New Session from the Data & AI Workspace Page

To start a new Data & AI Workspace session for an Embedded Data App, do the following:

  1. Open the Data & AI Workspace page.
  2. Select the Apps tab, and then select the Tetra Data App that you want to use. The configured Data App appears. Linked Data Apps appear in a separate tab in your browser, outside of the TDP. Embedded Data Apps appear within your workspace session within the TDP.

📘

NOTE

Embedded Data Apps based on Windows display a Session Files sidebar that shows all files that are part of your current workspace session. You can download files directly from the sidebar, or copy the file contents to your clipboard if the file is 15 MB or less.

Available Embedded Data App Instances

For Embedded Data Apps, each app's tile contains a number indicating how many instances are available to connect to. If no instances are available, this either means resources are scaling and one will be available shortly, or the maximum number of instances configured has been reached. To adjust the Maximum Users setting for Embedded Data Apps for your organization, see Edit Embedded Data Apps Settings for Your Org.

Start a New Session From the Search Page

📘

NOTE

Only Embedded Data Apps based on Windows support starting a workspace session from the Search page.

To start a new Data & AI Workspace session for an Embedded Data App based on Windows from the Search page, do the following:

  1. Sign in to the TDP. Then, in the left navigation menu, choose Search.
  2. On the Search page, filter the files that you want to use in your Data & AI Workspace session. For instructions, see Search .
  3. Next to Bulk Operations, select the Data & AI Workspace icon (this icon matches the Data & AI Workspace monitor icon from the left navigation menu).
  4. Select the Embedded Data App that you would like to open the files in. Then, select the Open [#] Filebutton. A new Data & AI Workspace session is created. It may take around a minute to open depending on the size of files selected.

📘

NOTE

The selected files are bundled together in a new data package that will be associated with your session and are loaded to the Documents\tetrascience_download folder.

Tetra FlowJo Data App

Open a Previous Data & AI Workspace Session

📘

NOTE

It is not possible to reconnect to an existing session by selecting additional files from the Search page. Any existing sessions will always be terminated before opening files in the Embedded Data App based on Windows.

To open a previous Data & AI Workspace session, do the following:

  1. Open the Data & AI Workspace page.
  2. Select the Sessions tab. A list of previous sessions appears.
  3. Find the session that you want to reopen. Then, select Open on the far right of that session's row. This will combine all of the previously selected files, and all of the analysis files uploaded back to TDP into a new data package to be used as part of your new session. If you click Open on an active session for your user, it will resume the existing session, not create a new one.

📘

NOTE

The Open button is deactivated in the following situations:

  • There is no available capacity for Embedded Data Apps.
  • The Embedded Data Apps has been deactivated, or it is in the process of updating.

Filtering Sessions

Sessions for all organization users are listed chronologically from newest to oldest. To help find the sessions you are looking for, the Sessions tab has two filters:

  • A User filter to limit the sessions to a particular Login User.
  • A Start date / End date filter to limit the sessions to a specific date range.

Session Information

Each session row displays the following information:

  • An application icon and name
  • An activity dot:
    • Grey means the session has ended
    • Green means the session is still running
  • When the session started
  • How long the session has been running for (or how long it ran for)
  • The user who started the session
  • The number of files the data package associated with the Data App session
  • The number of files uploaded back to TDP as part of the session

📘

NOTE

You can click on a session's # files downloaded or # files uploaded options to navigate to the Search page with filters pre-applied to show the specific files that were opened as part of the session.

Upload Files to the TDP From an Embedded Data App

To upload files from an Embedded Data App to the TDP, save the files at the root level of the Documents\tetrascience_upload folder in the Embedded Data App. The files are then uploaded to the TDP automatically within 15 seconds.

Keep in mind the following when uploading files from an Embedded Data App:

  • If no files from the TDP were previously opened in the Embedded Data App, the uploaded file will contain no inherited attributes .
  • The uploaded file's source type will indicate the type of Data App that the file originated from (for example, flowjo-data-app).
  • All uploaded files will have the following labels attached, which indicate the origin of the file and can help limit which data pipelines get triggered during upload.
    • ts_data_package_id: {data_package_id}
    • ts_data_analysis: true
    • ts_project_id: {project_id}
    • ts_project_name: {project_name}

📘

NOTE

Nested folder paths are currently not supported.