Permissions and Visibility
To roll out Conversation Analytics successfully, ensure the right people can:
- view transcripts/threads
- view AI insights and explanations
- access dashboards and saved searches
- (for admins) enable/override AI Tasks and manage Custom Fields
Recommended role model (typical)
A common approach:
- Tenant Admin: full configuration access (fields, tasks, overrides, dashboards)
- Supervisor / QA Lead: dashboards, advanced search, conversation details (read/write coaching notes if applicable)
- Analyst: dashboards and exports, limited configuration
- Agent / User: view assigned conversations, limited analytics
Field visibility
If your deployment supports field-level visibility: - keep sensitive insights restricted (e.g., churn risk, compliance flags, revenue metrics) - make operational insights broadly visible (e.g., topics, summaries)
MiaRec uses role-based permissions to control access to features. The specific roles and their permissions may vary by deployment.
Contact your provider if you need to adjust role permissions or restrict access to specific fields or dashboards.
EDITOR NOTE: RBAC and field visibility
Missing/unclear:
- Does MiaRec support field-level visibility?
- Are permissions per-feature, per-role, or both?
Ask product (choose one):
1) Permission model: - A) Fixed roles only - B) Custom roles with feature permissions - C) Both
2) Field visibility: - A) No field-level visibility controls - B) Field-level visibility by role - C) Field visibility by team/group
3) Dashboard permissions: - A) All users can see dashboards - B) Dashboards are permissioned by role/team - C) Dashboards can be shared manually
Best-guess recommendation:
- Provide at least role-based access for dashboards and sensitive fields.
- Document a minimal recommended setup (Admin + Supervisor + User).
Governance: who can change AI behavior
Prompt/filter overrides can change metrics and workflows. Decide:
- Who can enable/disable tasks?
- Who can override prompts/filters?
- Who can create custom insights?
Recommendation: limit prompt overrides to a small admin group and require testing before rollout.