Data residency
Understand how IBM Bob controls where your data is processed and stored across inference, data processing, and storage layers.
IBM Bob applies data residency controls across three independent layers: inference, data processing, and storage. When your organization selects a region, all three layers stay within that region. Some administrative data is maintained globally in US East regardless of your region selection.
Data residency layers
| Layer | Description |
|---|---|
| Inference | Model inference runs entirely within your selected region. |
| Data processing | Data pipelines that handle your content run only within your selected region. |
| Data storage | Conversation data is stored only within your selected region. |
Global data
The following data is stored in US East regardless of which region you select:
- Account and administrative data. Your IBM Bob account profile, organization settings, and admin dashboard configuration.
- Billing metadata. Bobcoin consumption records, subscription details, and invoice data that are required for billing and audit purposes.
- User and team management data. User identities, team memberships, role assignments, and enterprise policy configuration.
Available regions
IBM Bob operates in the following regions:
| Region | Location |
|---|---|
| US East | Washington D.C., USA |
| Europe | Frankfurt, Germany |
| India | Chennai, India |
| Japan | Tokyo, Japan |
Region selection is configured at the account or subscription level. Contact IBM to discuss region selection and restrictions for your organization.
When your organization moves from one region to another (for example, from US East to EU or Japan), account and administrative data remains in its global store and is not copied to the destination region. Conversation data is ephemeral and expires at session end. After migration, new conversations are processed in the destination region.