Guardrails and escalations are the governance controls that ensure AI Employees operate safely and accountably. They define the boundaries of AI action and the pathways for human oversight.
What Are Guardrails?
Guardrails are safety controls configured for each AI Employee that define:
- Scope Limits: What the AI can and cannot access
- Action Restrictions: Which operations require approval
- Data Boundaries: What information can be processed
- Compliance Rules: Industry-specific regulatory controls
Types of Guardrails
Opero supports multiple guardrail categories:
- Financial Limits: Caps on monetary values without approval
- PHI Protections: Healthcare data handling rules
- Communication Controls: Restrictions on external communications
- Tool Restrictions: Limits on which integrations can be used
- Time Boundaries: Operating hour restrictions
Escalation Workflows
When an AI Employee encounters a situation outside its guardrails, it automatically escalates to a human operator. The escalation includes full context, recommended actions, and the reason for escalation.
Configuring Guardrails
Guardrails are configured at the role level and can be customized per AI Employee. Organizations can set organization-wide policies that apply to all AI Employees, with role-specific overrides as needed.