Where your data goes
Understand the host, provider, and sharing choices behind the work.
Data handling depends on where the Project is hosted and which external services you choose.
Project records and execution
A Local Project’s files and work live on the local service. A Team Project’s shared records and execution live on its explicit host. Console supplies identity and organization-plan access; it receives no Project content through that role.
Assistant and prediction requests
A connected assistant provider receives the requests and context sent through that connection. An inference API receives the configured prediction requests, which can include Dataset samples. Inspect the provider and request mapping before using sensitive material.
Local inference executes on the service host with the installed engine and recorded license consent.
Private chat files
Attachments begin as private chat files. Save to Vault copies one into the Project; deleting the chat removes its private originals. Saved references retain exact record identities and revisions.
Diagnostics and sharing
Support reports provide a review step for included information before sending. Team membership and owner connection links grant different kinds of access; choose the appropriate sharing flow.
The product website’s privacy policy provides the maintained legal terms.