Skip to main content

Queue Management & Skill-Based Routing

Status: Planned

Overview​

Priority queues and skill-based routing for organizing requests and leads into named work queues, ensuring items reach the most qualified available agent. Required for government KCC deployments where multiple teams handle different inquiry types.

Standards​

  • GEMMA Klantgeleidingcomponent: gemmaonline.nl
  • TEC CRM: Section 3.2 (Assigning Cases), Section 3.3 (Escalating Unresolved Cases)
  • TEC CRM: Section 1.6 (Territory Management, Team Selling, and Member Reassignment)

Market Demand​

Queue management and skill-based routing are among the most requested capabilities in government CRM/KCC tooling, particularly for Dutch municipalities implementing the KCC-werkplek model (validated across 97K requirements from 39K+ tenders).

Key Capabilities​

Queue Entity​

  • Named, configurable queues (e.g., "Algemene Zaken", "Vergunningen", "Klachten")
  • Queue types: request queue, lead queue, mixed
  • Queue CRUD via admin settings
  • Queue membership: requests and leads can be placed in a queue

Priority-Based Ordering​

Items within a queue are automatically sorted by:

  1. Priority: urgent > high > normal > low
  2. Age: oldest item first within the same priority level

Skill-Based Routing (skill-routing)​

  • Skill tag system for Nextcloud users: admins tag agents with expertise areas (e.g., "vergunningen", "belastingen", "WMO")
  • Skill-to-category mapping: when a request enters a queue, the system evaluates available agents whose skills match the request category
  • Routing suggestion: the system suggests the best-matched, least-loaded agent
  • Auto-assign mode: optional automatic assignment without manual confirmation

Workload Balancing​

  • Agent workload indicator: current open items per agent
  • Round-robin fallback: when skill scores are equal, assign to the least-loaded agent

Queue Dashboard View​

  • Queue depth per queue (item count by priority)
  • Average wait time in queue
  • Agent workload table (open items per agent)

Data Model​

Two new schemas in pipelinq_register.json:

SchemaKey Fields
queuename, description, type (request/lead/mixed), isActive, routingMode
skillname, slug, description, category

Existing schemas modified:

  • request: gains queue (UUID reference, optional)
  • lead: gains queue (UUID reference, optional)
  • Nextcloud user profile: gains skills[] array (managed via admin settings)

Impact​

  • Data model: New queue and skill schemas; request and lead schemas gain queue field
  • Frontend: Queue list/detail views, queue dashboard widget, skill management in admin settings, queue column in request/lead list views
  • My Work: "My Work" view gains a queue-based section showing items from the agent's assigned queues
  • Admin Settings: Queue configuration and skill management sections

Specification​

Full specification: openspec/changes/archive/2026-03-22-queue-management/specs/

Related changes:

  • Design: openspec/changes/archive/2026-03-22-queue-management/design.md
  • Tasks: openspec/changes/archive/2026-03-22-queue-management/tasks.md