Skip to content
← Integrations

People & governance

The agent knows who to ask.

An escalation needs a destination, and a hard-coded name goes stale the first time somebody changes role. Connecting your workforce directory lets agents derive who currently holds a responsibility and who may approve what.

What connects

Your people, role and responsibility system, through its API.

What crosses the boundary

  • Into the agent: roles, teams, responsibilities and approval authority.
  • Out of the agent: nothing.
  • Never: salary, performance, assessment or absence data. Not exposed to any agent, on any plan.

The capabilities this exposes

  • Read team structure

Setting it up

  1. 1Authorise a read-only connection.
  2. 2Confirm the approval authority mapping.
  3. 3Grant team structure to the agents that escalate.

Permissions

Read-only, and limited to role, responsibility and approval authority. Salary, performance and assessment data is not requested and is not available to agents under any configuration.

Jobs an agent can own once this is connected

  • Escalate to whoever currently owns the responsibility
  • Route an approval to the person with the authority
  • Reassign an agent's owner when a role changes hands

Where it stops

Read-only. An agent cannot change a role, a responsibility or an approval authority. Deriving authority from here means it stays correct as people move — but it also means an out-of-date directory produces misrouted escalations, and the platform cannot tell the difference. This connection is not a route to monitoring your team. No product output scores an individual, and using this data to build a surveillance record breaches the acceptable use policy.

Give AI a job.

Build an agent, choose what it can access, define what it may do, and deploy it to work safely across your business.

No code required · Full SDKs and API · Stop any agent instantly · [email protected]