Researcher Workbench

Workspaces that were not manually migrated will be archived through a system‑led archival process. Archived workspaces are not active for analysis but may be recovered during the retention period. More information on this process will be provided soon.

Archived workspaces will be available within the Researcher Workbench at https://workbench.researchallofus.org/login. The process to retrieve an archived workspace is still under development, with more information coming soon.

GitRepo (Git repository) is intended for tracking and managing changes of code, scripts, and documentation, enabling version control. GitRepo is hosted outside of the Researcher Workbench and could be accessible outside of the Researcher Workbench by anyone with permissions to that repository. Please note that it would be a violation of the All of Us Data User Code of Conduct and Data and Statistics Dissemination Policy to share any participant-level data or direct participant counts of fewer than 20 outside of the Researcher Workbench. Therefore, to remain in compliance with All of Us policies, researchers should ensure that no participant-level data or counts of <20 (direct or inferable) are included in the code, scripts, or documentation included in the GitRepo when using GitRepo in their workspace. This includes in the outputs of the code or embedded into the code as comments or filters. Please be sure to go over your code and clear the outputs before including in GitRepo.
 

After you create a workspace, you can add a data collection to that workspace. You will add either the “All of Us Registered Tier” or “All of Us Controlled Tier” to your workspace via the Resources tab of your workspace. 

1. In the workspace, navigate to the Resource tab. Select Data from Catalog. 

2. Select the Data Collection of your choice. There will be two options: “All of Us Registered Tier” and “All of Us Controlled Tier.”

3. Select appropriate resources in the Data Collection. For example, for All of Us Registered Tier data, For example, for All of Us Registered Tier data, you will select R2025Q4R6, which is CDRv9 Registered Tier. A full list of CDR versions are noted in the Data Dictionary here. 

• CDR v9 Registered Tier = R2025Q4R6
• CDR v9 Controlled Tier = C2025Q4R6
• CDR v8 Registered Tier = R2024Q3R8
• CDR v8 Controlled Tier = C2024Q3R8
• CDR v7 Registered Tier = R2022Q4R13
• CDR v7 Controlled Tier = C2022Q4R13

4. Review the Data Collection policies, complete the “Researcher Use Statement Questions,” and select “I'm sure. I understand that all policies and terms above will be permanently applied to this workspace.”

5. You will have the option to select where in your resources you want this data collection to live. We recommend using your workspace folder path.
6. Select “Add to your workspace."

Data collections are curated datasets published in Verily Pre, the platform that powers the Researcher Workbench. There are currently two main All of Us data collections available in the new Researcher Workbench: “All of Us Registered Tier” and “All of Us Controlled Tier” which have multiple CDR versions. These data collections are synonymous with the curated datasets available in the legacy Researcher Workbench. 

When you log in to the Researcher Workbench using your @researchallofus.org username, you will automatically be provided access to any All of Us data collections for which you have completed the associated data access requirements. The same data access requirements that you are familiar with from the legacy Researcher Workbench (e.g., ID verification, Responsible Conduct of Research Training, Data User Code of Conduct attestation, etc.) are in place for gaining access to these data collections.

All data collections available in the Verily Pre platform, including the All of Us Research Program data collections, come with data collection policies that explicitly delineate built-in technical parameters to enforce data access and use restrictions. The same parameters are in place for All of Us data on the legacy Researcher Workbench. These parameters, for which Verily Pre broadly uses the term ‘policy,’ are distinct from the All of Us data access and use policies that you are already familiar with, which outline the program’s rules for access and use of All of Us data on the Researcher Workbench. These All of Us policies have not changed, and researchers are still responsible for reviewing and complying with them independent from the data collection policies.  The full list of data use and researcher policies can be viewed on this page.

Users can access both Registered Tier and Controlled Tier data in Researcher Workbench. However, older CDR versions (CDRv6, CDRv5, etc.) are not available. 

• CDR v9 Registered Tier = R2025Q4R6
• CDR v9 Controlled Tier = C2025Q4R6
• CDR v8 Registered Tier = R2024Q3R8 
• CDR v8 Controlled Tier = C2024Q3R8
• CDR v7 Registered Tier = R2022Q4R13
• CDR v7 Controlled Tier = C2022Q4R13

Researcher Workbench migration was completed June 2026. For more details regarding the migration timeline, please navigate to What to expect during the Researcher Workbench Migration.

To learn more about accessing the Researcher Researcher workbench, please see the comprehensive Researcher Workbench Getting Started Guide.

The Researcher Workbench, powered by Verily Pre, enables researchers to access, analyze, and collaborate on complex biomedical datasets with governance and policy oversight tools to safely support research speed and scale. 

The Researcher Workbench can be utilized by researchers at varying levels of computational expertise and skill through point-and-click interfaces and customizable programming tools:

• Analysis Applications
  • JupyterLab
  • JupyterLab Spark Cluster (aka Dataproc Cluster)
  • SAS
  • R analysis environment (aka RStudio) with customizable environment configuration (CPU, RAM)
  • Expanded cloud environment configurations (standard, high-memory and high-CPU machine types available)
  • Two NVIDIA environments - NVIDIA GPU‑accelerated libraries are designed for high‑performance, large‑scale computing environments. Please note: these environments are resource‑intensive and can incur significant computational costs. Be sure to review the associated expenses and confirm appropriate cloud resource allocations before provisioning these environments.
    • JupyterLab - NVIDIA NeMo for AI Development (NeMo)
    • JupyterLab - NVIDIA Parabricks and CUDA-X Data Science
• Workflows
  • UI and command line (CLI) support for Cromwell, Nexflow and dsub

To see more information on how to access your initial credits or set up a GCP billing account please visit the "Credits and Billing" section of the User Support Hub.

1. Navigate to the workspace you wish to add collaborator(s) to and then select the Share button.

2. Type the @researchallofus.org username of the collaborator(s) into the Email field. Then indicate the role you’d like to assign. The role determines the level of access and privileges they will have in the workspace. Select Done to apply the changes.

Note: You must be an Owner to share a workspace and view other collaborators within a workspace.