Usage Data in GoodData Cloud

GoodData Cloud collects usage statistics of the Product, which are analyzed in order to improve our services and provide the right functionality to our customers. Third-party tools Langfuse and Matomo are used for the collection and analysis of the data. GoodData Cloud collects information from the backend and frontend services and anonymizes/pseudonymizes to the maximum extent any personal information involved. In environments of all paying customers, user IDs are anonymized to protect individual identities. However, other metadata object identifiers, such as those for dashboards and data sources, are not anonymised, and we hereby don’t recommend including sensitive information as identifiers. The records collected and analyzed may typically include the following actions:

• Interactions with UI elements and APIs
• State of metadata and configuration of individual organizations
• Conversations and interactions with AI-driven elements of our interface (such as search, chatbot, …)

See the GoodData Privacy Policy for more information on how this data is used by GoodData as well as your rights and GoodData’s obligations with respect to this data.

AI Assistant Troubleshooting Data Retention

GoodData stores AI Assistant conversation data for troubleshooting and observability. Usage and troubleshooting data is anonymized or pseudonymized to the maximum extent possible, as described above. By default, AI Assistant conversation data is retained for 365 days.

Authorized administrators can configure the retention period through the API by using the conversationRetentionDays setting. The setting can be configured at the organization or workspace level. A workspace-specific value overrides an inherited value. If no value is configured, GoodData uses the default retention period.

Configure Retention Through the API

Use the existing settings API to configure the retention period:

• For an organization-level value, use POST /api/v1/entities/organizationSettings and set the setting type to conversationRetentionDays.
• To override the organization-level value for a workspace, use POST /api/v1/entities/workspaces/<WORKSPACE_ID>/workspaceSettings and set the setting type to conversationRetentionDays.
• To check the effective value for a workspace, use GET /api/v1/actions/workspaces/<WORKSPACE_ID>/resolveSettings, find the setting with type conversationRetentionDays, and read its value from content.

The value is an integer number of days. GoodData validates submitted values against the currently supported range (minimum 1 day, maximum 365 days) and rejects values outside that range.

Changing the retention period does not change what troubleshooting data is captured or how it is masked. It only controls how long the stored data is kept.

Retention is calculated from the conversation’s last activity, not from when the conversation was created. Pinned conversations are subject to the same retention policy.

Expired conversations are purged automatically by a daily cleanup process. Because cleanup runs periodically, data may remain available for some time after it reaches the configured retention threshold.

A shorter retention period can help meet stricter data-minimization or compliance requirements, but it also reduces how far back GoodData can investigate past issues. When choosing a retention period, consider the balance between your compliance requirements and the troubleshooting history you want to keep available.

For information about auditing changes to the AI data-sharing configuration, see the Auditability section.