How authentication works
Rippit connects to your helpdesk through the OAuth 2.0 authorization code grant. You authenticate against your helpdesk, never against Rippit, and Rippit never receives a password or an API key. The authorization request carries a single-use state nonce that expires if it is not redeemed, so a replayed callback matches nothing. After the exchange, Rippit verifies the token against the source system before storing the connection, and rejects a source account already connected to your workspace. Where the source system supports a scope parameter, Rippit requests read access in this step. Zendesk receivesscope=read. Intercom has no scope parameter, so permissions come from the Intercom app configuration shown on its authorization screen.
The token Rippit holds
Rippit stores a long-lived access token for the account that authorized the connection, and uses it for every scheduled sync. It acts as that account, which is why the account you choose determines what Rippit can reach. Revoking the grant in your helpdesk invalidates the token immediately.Access scope
Rippit issues no writes. There is no code path that creates, updates, or deletes anything in your helpdesk. Every Zendesk call is aGET, and Rippit creates no Intercom conversations, replies, notes, or tags.
Rippit can read only what the authorizing account can read. Use a dedicated service account where the source system supports one, so you control what Rippit can see.
What reaches Rippit
Full conversation content, not just metadata: message bodies, participants, and attached fields including custom fields. Zendesk attachments are downloaded and stored rather than linked. Rippit does not train any AI models. Your data is used for inference only.Revoking access
Revoking Rippit’s authorization in the source system stops the token working immediately. Disconnect in Rippit removes the stored connection and stops future syncing. Neither deletes conversations already imported. See how data integrations work for what account deletion does and does not remove.Hosting and compliance
Subprocessors: AWS (US and EU), MongoDB, Coralogix, Fullstory (no end user data), Google Cloud Platform (Gemini, US only, where used).
Signing up for Rippit accepts the Terms of Service and Privacy Policy. No separately negotiated contract is required, and the Privacy Policy is what governs personal data by default.
A standard Data Processing Addendum is available on request, alongside current certificates, audit reports, and the full subprocessor list, from the Rippit Trust Center.