Skip to main content

Data Storage & Security

  • 20 August 2024
  • 0 replies
  • 25 views

Data Storage

AI Studio does not store any data, it simply identifies where the data exists and provide tools for data processing. Users have complete control of their own data governance and systems. 

 

Data Source Ingestion

Linking to the Data Source: Users can link AI Studio to a data source like an S3 bucket by providing the necessary pointers or connection details. This step establishes a connection between AI Studio and the data stored in the cloud (in this case, AWS).

 

Downloading Data: Depending on specific project needs and the setup of AI Studio, users can download the required data from the source (like the S3 bucket) to their local device.

 

Data Stays on the User's Device: Once downloaded, the data resides on the user’s local device.

 

Data Governance and Security: users are responsible for managing the access, security, and compliance aspects of the data used with AI Studio.

 

AI Studio Access

Local: AI Studio allows users to access and manage data stored directly on their local device.

 

Cloud: In the early access phase, HP will support AWS (Amazon Web Services) only.

 

In the future, HP plans to extend support to other major cloud platforms like Azure (Microsoft's cloud service) and Google Cloud as well.

 

With AI Studio, users can access their data from any geographical location as long as they have the necessary setup. This means if the user has AI Studio installed on their local device and an internet connection, they can access local data from wherever they are. For cloud-based data, geographical location doesn't matter as long as the user can connect to the internet and access the cloud service (like AWS).

 

Authentication Policies and Tools

Application data (projects, workspaces, datasets) are synchronized with a cloud database whenever there is internet connectivity.

 

Files (either shared files, mlflow artifacts, tensorboard logs) are synchronized peer-to-peer with end-to-end encryption.

 

AI Studio uses a cloud relay in the situations where a direct peer-to-peer connection cannot be established (but the encryption is still between the peers themselves, cloud only sees ciphertext).

 

Be the first to reply!

Reply