Currently, OCR capacity rules are defined per channel and applied per agent, meaning an agent's email capacity is a single shared counter across all groups they belong to. This creates a significant routing problem for teams that operate across multiple groups with different ticket types and priority structures.
Our specific case: we have agents who belong to both a X Emails group and a X Alerts group. When an agent has 3 assigned email tickets in the X Emails group, OCR considers them at capacity for email and will not push tickets from the X Alerts group — even though alerts are higher priority and the agent has the bandwidth to handle them in parallel. The result is that lower-priority assigned tickets block the routing of higher-priority unassigned ones, which directly causes SLA breaches on our most urgent work.
The requested change is the ability to define capacity limits at the group level, so that an agent's capacity in Group A is tracked independently from their capacity in Group B. For example: an agent could have a maximum of 3 open email tickets in X Emails and separately a maximum of 3 open email tickets in X Alerts, with OCR evaluating each group's queue against only that group's capacity counter.
This would unlock accurate push-based routing for teams with multi-group structures and tiered priority queues — a common pattern for operations, trust and safety, and regulatory/compliance/risk teams — without requiring custom middleware or workarounds that reintroduce complexity OCR is designed to eliminate.