Direct Answer (TL;DR)
Brilo AI Edit Restrictions control who can change a Brilo AI voice agent’s behavior, scripts, voice settings, or response logic. By default, only users with workspace-level administrative or edit permissions can update agent scripts, adjust voice or SSML settings, and change routing or escalation rules; more granular role controls can limit who publishes changes to live agents. Use edit restrictions to separate draft editing from production publishing, require approvals for custom voices or SSML, and reduce accidental changes to agent behavior.
Who can modify Brilo AI voice agent behavior or scripts?
Who can edit agent scripts? Only workspace admins or users granted edit permissions; publishing to production may require an approval step.
Who can change voice or SSML? Admins or designated voice editors; custom voice requests often require a support request.
Who can update routing/escalation? Users with routing or admin permissions; handoff rules typically require workspace-level access.
Why This Question Comes Up (problem context)
Enterprises ask about Edit Restrictions because agent behavior affects compliance, brand voice, and customer experience. In regulated sectors such as healthcare, banking, and insurance, a single unauthorized change can expose sensitive workflows or break escalation rules. Buyers need clarity on who can make edits, how changes are tested, and how Brilo AI supports separation between authors, testers, and production operators.
How It Works (High-Level)
Brilo AI enforces Edit Restrictions through workspace roles and agent-level permissions. Typical workflow: an editor creates or updates a draft agent script, a reviewer runs controlled test calls, and an admin publishes approved changes to the live agent. Brilo AI stores agent configurations (voice, SSML, routing, and answer length) in the workspace so changes are auditable and can be rolled back when needed.
Edit Restrictions limit which users can change or publish agent settings and scripts. An agent script is the set of prompts, prompt order, and response rules that define how a Brilo AI voice agent speaks and responds.
Related setup guidance for agent settings and test workflows is available in the Brilo AI agent configuration guide: Brilo AI agent settings and long-conversation setup.
Technical terms used in this article include edit permissions, role-based access control (RBAC), SSML (speech markup), agent settings, routing, and publish/rollback.
Guardrails & Boundaries
Do not allow unrestricted publishing. Limit publish rights to a small group of workspace admins to prevent accidental live changes.
Require testing before publish. Use a test agent or sandbox calls to validate script and voice changes.
Restrict advanced voice edits. Requests for custom prosody, voice cloning, or SSML changes may require a formal support request and legal consent workflows before they can be enabled.
Workspace admin is the role that can change workspace-level settings, grant user permissions, and approve or publish agent changes.
Brilo AI will not automatically push untested drafts to production, will not bypass configured escalation rules, and will not change recording or data-retention settings without explicit admin configuration. If a change could affect recording, transcription, or personal data handling, Brilo AI requires confirmation of workspace policies before long-term storage is enabled. For details on support-required voice capabilities and SSML guidance, see the Brilo AI SSML and voice tuning guidance: Brilo AI SSML and naturalness guidance.
Applied Examples
Healthcare: A clinical call center restricts editing so only a compliance reviewer and a clinical content owner can publish changes to scripts that collect clinical triage information. The Brilo AI voice agent draft is tested in a sandbox with recordings disabled until legal confirms the wording is compliant.
Banking: A retail bank separates duties so product teams can author scripts, but only system admins can modify routing rules that trigger secure verification or escalate to fraud teams. The bank uses role-based approvals to prevent unauthorized changes to authentication prompts.
Insurance: An insurer requires that any SSML or custom voice request go through support and legal review before deployment; changes to callback scheduling or hold messaging must be validated in test calls.
Human Handoff & Escalation
Brilo AI voice agent workflows can be configured to hand off to a human or another workflow when an edit or script path reaches a boundary. Common handoff triggers include low confidence in intent detection, a customer request for a human, or a configured escalation rule. When configured, the Brilo AI voice agent passes caller context, recent transcript snippets, and the agent state to the human agent or downstream system so the human does not need to repeat collection steps. Edit Restrictions ensure only authorized users can change those handoff rules so escalation behavior remains predictable.
Setup Requirements
Grant workspace-level access: Provide Brilo AI with a list of workspace admins and editors to seed permissions.
Define roles: Create role definitions (publishers, editors, reviewers) so responsibilities and publish rights are clear.
Configure agent-level permissions: Set which roles can edit drafts, run test calls, and publish live agents in the Brilo AI workspace.
Upload test artifacts: Provide sample call flows, test numbers, and any required audio samples for voice tuning.
Enable testing: Run controlled test calls in a sandbox agent and log results before approving changes.
Approve and publish: Have reviewers sign off and have an admin publish the change to production.
Monitor and rollback: After publishing, monitor agent performance and use Brilo AI’s rollback option if unexpected behavior occurs.
For detailed agent-level configuration and test workflows, see the Brilo AI agent settings and long-conversation setup guide: Brilo AI agent settings and long-conversation setup.
Business Outcomes
Proper Edit Restrictions reduce risk, improve compliance posture, and maintain a consistent brand voice. Operational benefits include fewer production incidents from accidental edits, clearer audit trails for who changed scripts and why, and faster, safer rollouts because changes pass through defined test and approval gates. These controls help regulated teams maintain separation of duties while enabling iterative improvements to Brilo AI voice agent behavior.
FAQs
Who can publish a script change to a live Brilo AI voice agent?
Publishing rights are typically limited to workspace admins or a small group of designated publishers. Authors can create and test drafts but may need an admin to approve and publish to production.
Can I allow non-admins to edit scripts but not publish them?
Yes. Brilo AI supports role separation where editors can change draft scripts and testers can run sandbox calls, but only approvers or admins can publish to the live agent.
What happens if someone publishes a breaking change?
Brilo AI recommends rolling back to the previous agent version and running test calls; workspace admins should monitor logs and recordings to diagnose the change. Maintain a small publish group to reduce this risk.
Are custom voices and SSML edits allowed directly in the dashboard?
Basic SSML and voice options are available in agent settings, but custom prosody or voice cloning requests usually require a Brilo AI support request and any required approvals before enabling in production.
How are edits audited?
Brilo AI records who made changes to agent settings and when; combine this with controlled publishing to create an auditable change history for compliance reviews.
Next Step
Review and apply role best practices in your workspace by checking the Brilo AI agent settings and long-conversation setup guide: Brilo AI agent settings and long-conversation setup.
Validate voice and SSML change requirements using Brilo AI’s naturalness and SSML guidance: Brilo AI SSML and naturalness guidance.
If you need a step-by-step authoring and approval workflow, follow the dashboard agent configuration and testing instructions: Brilo AI agent settings and voice configuration guide.