Clarino

Security

Last updated October 27, 2026

This page describes the controls Clarino uses to protect your workspace, the people you talk to, and the systems you connect.

Accounts and access

Sign-in is handled by a dedicated identity provider with support for organizations. Team members join a workspace by invitation and receive a role: owner, admin, member, or viewer. Only owners and admins can change who is on the team, publish the voice agent, change billing, or adjust data controls. Viewers are read-only.

Workspace isolation

Every record Clarino stores is tagged with the workspace it belongs to, and the database enforces that boundary with row-level security on every query. Vector knowledge used for answers is stored in a separate namespace per workspace. A workspace cannot read, search, or receive another workspace's data.

Integrations and webhooks

Events from the voice provider are accepted only when they carry a valid signature over the exact request body, and they are matched to a workspace through records Clarino already owns, never through identifiers supplied in the request. Events Clarino forwards to your own endpoints are signed with a per-workspace secret and carry a delivery ID and timestamp so you can verify and deduplicate them.

Payments

Billing runs through a PCI-compliant payment processor. Clarino never sees or stores card numbers. Billing events are verified against a signed webhook secret before they change a workspace's plan.

Data controls

Admins can turn on redaction of sensitive details such as card numbers and personal data before transcripts are stored, and choose how long conversations and calls are kept: 30 days, 90 days, or one year. Expired content is removed by a daily job that leaves an audit record. Workspace data can be exported from Settings at any time.

Transport and storage

All traffic to Clarino uses HTTPS. Workspace data is stored with a managed database provider that encrypts data at rest. Secrets used to reach providers live only in the production environment and are never written to the codebase.

Reporting a concern

If you believe you have found a security issue, email support@clarino.ai with the details and we will respond as quickly as we can. Please do not test against other customers' workspaces.