Data retention policy
We retain only the minimum data required to operate ApproveFlow.
Expense request records (amount, description, status, timestamps, user IDs) are stored for the purpose of tracking request history and usage per workspace.
We do not store any Slack messages, user profile information, or channel content.
All data is retained for as long as the workspace is actively using the app. Upon uninstalling ApproveFlow, all associated data is automatically deleted within 30 days.
Workspace administrators can request immediate deletion of data at any time by contacting us at support@approveflow.io.
Data archiving and removal policy
When a workspace uninstalls ApproveFlow, all related data (expense requests, user references, and metadata) is flagged for deletion.
We automatically delete all workspace-related data within 30 days of uninstallation.
We do not archive any data after removal.
Workspace administrators may also request immediate deletion of all data by emailing support@approveflow.io.
Data storage policy
All data is stored securely on AWS infrastructure within the European Union (EU), using encrypted PostgreSQL databases.
All connections to our services are secured via HTTPS/TLS. We do not store any Slack message content or credentials.
Only authorized backend services and a limited number of approved team members have access to the production database, protected by role-based access controls and audit logging.
We do not replicate data outside of the EU.
Data center location(s)
Germany
Data hosting details
We host all application and database services on AWS, using virtual private cloud infrastructure with strict access controls.
The application runs in Docker containers, and data is stored in encrypted PostgreSQL databases.
All traffic is secured via HTTPS, and data access is limited to backend services and authorized personnel only.
All data is stored in AWS data centers located in the European Union. Primary region: Frankfurt, Germany (eu-central-1).
App/service has sub-processors
yes
Guidelines for sub-processors
App/service uses large language models (LLM)
no