PATT PROPATT PRO · privacy
How this connection handles information.
This service-specific notice describes the PATT PRO AI connection provided under licence by SIYA CONECTA. Contact support@patt.club for privacy questions or requests.
The connection lets you request read-only PATT information through ChatGPT within your current permissions. This notice concerns the PATT PRO AI connection and the information it processes.
Information processed
| Purpose and access | Information involved |
|---|---|
| Sign-in and connection security | Your PATT email/username and password are processed on PATT to verify the account. Account identity, permissions, credential-revision checks, approved scopes and authorization/session records support continuing access. Passwords and password hashes are not returned to ChatGPT by this flow. |
| Event and staff reporting | Permitted event, ticket and table configuration; operational counts and monetary estimates; permitted staff/team labels; and event-entry or booking records without client names or contacts. The access-context result includes the connected account’s ID, type and name. |
| Additional client access | Permitted client names, activity and selected history, count/date filters and operational-value summaries. CRM authorization applies independently of reporting shares. |
| Additional contact access | A selected authorized client’s email and phone when available. There is no bulk contact-export action. |
| Additional birthday access | Stored birthday data is processed for explicitly requested month/day queries. Birthday matches and source coverage, or an opted-in selected profile’s month/day, can be returned. Birth years are not returned. Stored values may be unverified or generated defaults. |
Consent to connect does not expand your existing PATT access, establish marketing consent or authorize disclosure of unrelated people’s information. Opaque or pseudonymous record references may still be identifiable when combined with other information.
Recipients and purposes
Requested results are sent to the ChatGPT connection you authorize. The AI provider’s policies, contracts and your account settings govern its handling of those results. PATT uses the data described here to authenticate the account, enforce approved access, answer requested queries and maintain service security and audit records.
The connection uses Azure for service hosting and authorization infrastructure, and MongoDB for source business data. The current connection server is hosted in Azure France Central; this does not establish where every database copy, backup or AI-provider record is stored.
Sessions, continuing access and storage
Essential session cookies and stored authorization state allow the connection to work across requests and restarts. Authorization payloads are encrypted. Limited technical audit records include tool names, outcomes, timing and pseudonymous identifiers; reviewed audit outputs exclude clear contact values and tool-result contents.
| State | Current technical behavior |
|---|---|
| Access and sign-in validity | Access/ID tokens and provider/management sessions have 15-minute validity. Authorization codes have 60 seconds; authorization interaction state has 10 minutes. |
| Continuing access | Grants last up to 30 days. Refresh tokens have a configured seven-day lifetime and rotate. Current account permissions, grant validity and revocation checks continue to apply. |
| Expired authorization payloads | Expired encrypted payloads are pruned in batches during store writes. An idle store can retain expired bytes beyond their authority expiry; expiry is not a guaranteed deletion deadline. |
| Revocation metadata | Keyed identifier hashes and denial metadata have no time-expiry policy, to prevent stale operations restoring revoked access. |
| Business results | The reviewed reader processes results in memory without a durable result cache. Source PATT records and copies received by ChatGPT have separate retention. |
Operational retention and requests
Limited operational and security records are retained for the purposes for which they are needed: operating and protecting the connection, preventing abuse and replay, investigating incidents, handling requests and meeting applicable obligations. This includes retained revocation metadata needed to prevent old authorization state restoring access. Any service backups are subject to those purposes and can retain earlier records until they are replaced or removed.
We do not promise a fixed automatic deletion date for technical logs or backups. Token expiry, browser sign-out and disconnect are access controls, not a promise that all stored records are immediately erased. A request can be made to the monitored privacy contact below; retention needed for security or applicable obligations can affect what can be removed.
SIYA CONECTA handles requests about connection and authorization data through support@patt.club. Requests involving original event, attendee or staff records are coordinated with the relevant organiser or account responsible for those records. The connection itself does not edit or erase those business records.
Disconnect and privacy requests
Use Disconnect app in Connected AI apps to revoke the connection. Signing out of that page alone keeps app grants active. Disconnecting does not delete the original PATT records, previously delivered ChatGPT results or files exported by a user.
Send privacy questions or requests to support@patt.club. Describe the request without sending passwords, tokens or unnecessary customer records. Copies already held by the AI provider are also subject to its own controls and policies.