Applied LabsDocs
Skip to Content

Routing and Assignment

Routing automatically assigns escalated conversations to the right human agent so your team can respond quickly without manually triaging every incoming request.

Groups

Groups are teams of agents that you create based on your internal team structure. For example, if you have 2-3 agents who should always handle fundraising questions, you can create a group for them and configure routing rules so that topic is assigned only to that group. Groups are used throughout the Help Desk for assignment, capacity management, and escalation targeting.

How routing works

When a conversation is escalated from the AI to a human agent, the system immediately runs through a series of rules to determine who should handle it.

  1. Mapping rules are evaluated first. Each mapping rule matches conversation attributes — such as topic or intent — to a target agent group. Rules are evaluated in the order you define them. If a rule matches, the system uses that group’s members as the candidate pool for assignment. You can mark a rule as stop on match to prevent any further rules from being evaluated once it matches.

  2. If no mapping rule matches, the system falls back to round-robin assignment across all eligible members in your workspace.

  3. Capacity rules are applied to the candidate pool. Before assigning a conversation, the system checks whether each candidate agent has room to take it on. Capacity rules are configured per group and define two things: the maximum number of concurrent escalated conversations an agent can hold, and a timeout window — any conversation without a recent message within that window does not count against the agent’s limit.

  4. Per-platform capacity keys let you manage channel loads independently. For example, you can set a separate limit for Facebook chat versus Instagram comments, so a surge on one channel does not consume slots meant for another.

Priority rules

Priority rules let you configure how urgently a conversation is treated based on its attributes. You can set priority based on:

  • Channel — for example, treat phone calls as higher priority than email.
  • Contact group — prioritize VIP customers or specific segments.
  • Contact context — use arbitrary key-value data attached to the contact.
  • Topics and labels — escalate conversations about billing or outages ahead of general inquiries.

Priority affects how conversations surface in the queue.

Agent presence and status

The system uses each agent’s presence status to determine whether they are eligible to receive new conversations. There are four statuses:

  • Active — the agent is signed in and available. If the agent stops interacting with the app, they are set to offline after 15 minutes.
  • Away — the agent is signed in but temporarily unavailable. This status is set manually and persists until the agent changes it.
  • Offline — the agent is not available..
  • Out of Office — the agent is on extended leave. Agents with this status are skipped entirely during round-robin assignment.

Manual statuses (Away, Offline, Out of Office) are sticky — they do not expire automatically and remain until the agent updates them.

Agents who are Out of Office are never assigned conversations, regardless of routing rules. You can also configure routing to only assign to agents whose status is Active.

If an agent is currently on a phone call, the system will not route new phone or chat conversations to them until the call ends.

Next steps

Last updated on
Keyboard ShortcutsSLA Policies