Skip to main content
If your firm signs client security questionnaires, you already know the questions: where does our data go, who can see it, and what record is kept of who did what? Parlé is architected to give you defensible answers — single sign-on tied to your identity provider, project files stored inside your own SharePoint or Google Drive tenant, an append-only audit log enforced at the platform level, and role-based access that limits who can see what down to the project level.

Single Sign-On

Connect Microsoft Entra ID or Google Workspace. Offboarding a user in your identity provider immediately removes their Parlé access — no manual deprovisioning required.

Native Storage

On the Business plan and above, project files live in your own SharePoint site or Google Drive tenant. Client documents never leave your environment.

Audit Log

Every sensitive action is recorded in an append-only log. Parlé has no feature that edits or deletes an entry, and the database rejects ordinary updates and deletes. On the Business plan and above, Org Admins can read, filter and export their organization’s log in Org settings.

Role-Based Access

Four roles span org-level and project-level control. A Restricted project is visible only to its members and to your Org Admins; everyone else cannot see that it exists.

​
What Parlé Does Not Do

Understanding the boundaries is just as important as understanding what is included. Parlé does not:
  • Use personal consumer AI accounts. All model calls run on your own provider API keys, or a managed Parlé model — never through a personal OpenAI, Anthropic, or Google account.
  • Store files outside your tenant (Business and above). On the Business plan and above, uploaded project files are written directly to your SharePoint or Google Drive. Parlé processes them in transit to the model but does not retain them in Parlé-managed storage.
  • Mix session data across organizations. Every organization is fully isolated. Conversations, files, members, and usage data are scoped to your org and never visible to another org on the platform.

​
Answering the AI Security Questionnaire

Use the answers below when completing client or internal security questionnaires. Each answer names the plan it applies to where that varies by plan.
On the Business plan and above, project files are stored in your own SharePoint site or Google Drive tenant — the same environment your firm already governs. Parlé processes file content in transit when sending it to the AI model, but does not persistently store that content outside your tenant. On the Team plan, files are stored in Parlé-managed secure storage.
The provider you choose and connect via your own API key. Parlé supports Anthropic, OpenAI, Google Gemini, and Azure AI Foundry — bringing your own Azure AI Foundry deployment requires the Business Elite plan and above (managed models are unaffected). You control the provider relationship, the key, and — where your provider agreement permits — the data processing terms. Parlé never routes your requests through a shared or pooled consumer account.
It depends on the project’s visibility. A conversation in a Restricted project can be opened only by that project’s members and by your Org Admins, who can open every project in the organization. A conversation in a Public project can be opened by anyone in your organization. A personal conversation is visible only to the person who owns it. SSO-offboarded users have their session invalidated and access removed immediately when their account is disabled or removed in your identity provider.
No. There is no view in Parlé that allows cross-organization access to your conversations. On all plans, conversations are scoped to your organization and are not accessible to Parlé staff in the normal course of operations.
For the complete security and compliance details, visit parlehub.com/security.