Support AI Skill

Intelligent Ticket Routing

Route tickets to the most qualified agent based on expertise, workload, availability, and historical performance. Use when configuring skill-based routing, setting up workload balancing, implementing language matching, or building follow-the-sun routing for global teams. Triggers on phrases like 'ticket routing', 'agent assignment', 'skill-based routing', 'workload balancing', 'auto-assign tickets', 'follow-the-sun', 'language matching', 'routing algorithm', 'queue management'.

Intelligent Ticket Routing & Agent Assignment

Route every ticket to the optimal agent using a multi-factor algorithm that balances skill match, workload, availability, historical performance, and customer preferences.

Workflow

Phase 1: Routing Configuration

Define agent skill profiles:

Technical competencies (database, API, frontend, mobile, integrations)
Product expertise (by product line, module, version)
Language fluency (primary, secondary, certification level)
Customer tier authorization (enterprise-only agents, VIP handlers)
Soft skills (de-escalation, executive communication, training)

Set routing algorithm parameters:

Primary: skill match (40%)
Secondary: workload balance (25%)
Tertiary: historical performance with similar issues (20%)
Quaternary: customer preference/agent relationship (10%)
Tertiary constraint: time zone alignment (5%)

Configure queue hierarchy:

Primary queue → secondary queue → overflow queue → manual pool
Backstop routing when no agent matches (team lead assignment)
Auto-queue when all agents at capacity (wait time estimation)

Phase 2: Real-Time Routing Engine

Incoming ticket evaluation:

Extract category, priority, language, customer tier from ticket metadata
Identify required skill set from category mapping
Check for customer-specified agent preference

Agent availability scan:

Query online agents with matching skills
Filter by current workload (active ticket count vs. capacity)
Filter by availability status (online, away, DND, offline)
Check time zone alignment for expected follow-up windows

Score & rank candidates:

Calculate composite routing score for each eligible agent
Rank by score; assign to top candidate
If no match within 30 seconds: escalate to team lead pool

Assignment & notification:

Assign ticket with routing reason code
Push notification to agent (Slack, email, desktop, mobile)
Start SLA timer; update queue dashboard

Phase 3: Dynamic Re-routing & Load Management

Automatic re-routing triggers:

Agent goes offline/away → reassign to next eligible agent
SLA breach risk (no activity for 50% of SLA) → reassign
Agent workload exceeds threshold → redistribute new tickets
Customer explicitly requests different agent

Overflow management:

Queue depth monitoring with capacity alerts
Auto-scale: wake on-call agents when queue > threshold
Overflow routing to secondary team or general pool
Customer communication: estimated wait time updates

Performance tracking:

Track routing accuracy (did first assignment resolve?)
Monitor re-routing rate (target: <15%)
Agent satisfaction with routing fairness

Templates

Routing Algorithm Configuration

INTELLIGENT ROUTING CONFIGURATION
===================================
Version: [4.0] | Last Updated: [Date]

ROUTING FACTORS & WEIGHTS:
┌─────────────────────────────┬────────┬──────────────────────────────┐
│ Factor                      │ Weight │ Configuration                │
├─────────────────────────────┼────────┼──────────────────────────────┤
│ Skill Match                 │ 40%    │ Exact match = 40,            │
│                             │        │ Partial match = 25,          │
│                             │        │ No match = 0                 │
│ Current Workload            │ 25%    │ 0-2 active = 25 points,      │
│                             │        │ 3-5 active = 15 points,      │
│                             │        │ 6+ active = 5 points         │
│ Historical Performance      │ 20%    │ FCR rate with similar issues │
│                             │        │ (>80% = 20, 60-80% = 12,    │
│                             │        │ <60% = 5)                    │
│ Customer Relationship       │ 10%    │ Same agent handled before    │
│                             │        │ (+10 if >3 prior tickets)    │
│ Time Zone Alignment         │ 5%     │ Same timezone = 5 points,    │
│                             │        │ Within 2 hours = 3 points    │
└─────────────────────────────┴────────┴──────────────────────────────┘

QUEUE HIERARCHY:
  Level 1: Specialized team queue (exact skill match)
  Level 2: General team queue (partial skill match)
  Level 3: Overflow pool (any available agent)
  Level 4: Manual assignment (team lead decision)
  Backstop: Auto-assign to least-loaded agent (any category)

CAPACITY RULES:
  Max concurrent tickets per agent: 8
  Max P0/P1 tickets per agent: 2
  Cool-down period after P0 assignment: 30 minutes
  Away status: pause new assignments for 15 minutes
  Offline: reassign all active tickets within 10 minutes

RE-ROUTING TRIGGERS:
  - No agent activity > 50% of SLA time → reassign
  - Agent status change (online → offline) → immediate reassign
  - Workload exceeds 1.5x capacity → pause new routing to agent
  - Customer requests different agent → manual override

Agent Skill Profile Template

AGENT SKILL PROFILE — [Agent Name]
====================================
Employee ID: [ID] | Team: [Team Name] | Start Date: [Date]

TECHNICAL COMPETENCIES:
  Database:        ██████████░░ Expert (9/10)
  API/Integration: ████████░░░░ Advanced (7/10)
  Frontend/UI:     ██████░░░░░░ Intermediate (5/10)
  Mobile Apps:     ████░░░░░░░░ Basic (3/10)
  Security:        ███████░░░░░ Proficient (6/10)

PRODUCT EXPERTISE:
  Product A (Core Platform): Certified ✓
  Product B (Analytics):     Proficient
  Product C (Mobile):        Basic
  Product D (Enterprise):    Certified ✓

LANGUAGES:
  English: Native
  Spanish: Fluent (certified)
  French:  Conversational

AUTHORIZATIONS:
  Enterprise tier support: ✓
  VIP/Executive handling:  ✓
  Security incident response: ✗ (training pending)
  Refund processing (>$1000): ✗ (manager approval required)

PERFORMANCE METRICS (Last 90 Days):
  First Contact Resolution: 78% (team avg: 72%)
  Average Handle Time: 12.3 min (team avg: 14.1 min)
  CSAT Score: 4.6/5.0 (team avg: 4.3/5.0)
  Tickets Resolved: 342 (team avg: 298)
  Routing Accuracy: 91% (assigned → resolved without re-route)

Integration Points

Ticketing systems: Zendesk, Freshdesk, Intercom, Jira Service Desk, Front
CRM: Salesforce, HubSpot (customer tier, history, preferences)
WFM (Workforce Management): NICE, Verint, Calabrio (scheduling, capacity)
Communication: Slack, Microsoft Teams, Twilio (agent notifications)
Analytics: Tableau, Power BI (routing dashboards, agent scorecards)
Agent desktop: Zendesk Agent Workspace, Front Inbox
Timezone services: World Time API, holiday calendar integrations
HR systems: Workday, BambooHR (agent skill profile data)

Edge Cases

| Scenario | Handling | |----------|----------| | All qualified agents offline | Route to overflow pool; notify team lead; extend SLA with customer notification | | Agent capacity exceeded team-wide | Activate overflow routing; page on-call agents; auto-reply with wait estimate | | Customer requests unavailable agent | Assign to backup agent; notify customer; offer callback option | | Multi-language ticket | Language detection → route to fluent agent; if none, use translation layer | | Agent on PTO/unexpected absence | Pre-assignment redistribution; auto-reassign active tickets within 15 min | | VIP customer + no specialist available | Escalate to team lead; bypass skill matching for authority-based routing | | Ticket requires cross-team collaboration | Assign to primary owner; auto-cc secondary team; create linked tickets | | Round-robin vs skill-based conflict | Skill-based takes priority; round-robin only among equal-scored candidates |

Output

Routing Dashboard

ROUTING PERFORMANCE — Real-Time Dashboard
==========================================
As of: 2025-01-15 10:30 UTC

ROUTING METRICS:
  Tickets routed today: 347
  Auto-assigned: 318 (91.6%)
  Manual assignment: 22 (6.3%)
  Overflow pool: 7 (2.0%)
  Average routing time: 1.4 seconds

AGENT UTILIZATION:
  Online agents: 28 / 35 total
  Average active tickets/agent: 4.2 (capacity: 8)
  Highest workload: Agent Sarah (7/8 — approaching limit)
  Lowest workload: Agent Tom (1/8 — eligible for overflow)

QUEUE DEPTH:
  Engineering queue:    12 tickets (avg wait: 2 min)
  Billing queue:         5 tickets (avg wait: <1 min)
  General Support:      18 tickets (avg wait: 4 min)
  Overflow pool:         3 tickets (avg wait: 8 min)

ROUTING QUALITY (Last 7 Days):
  First-assignment resolution rate: 73.4%
  Re-routing rate: 11.2% (Target: <15%)
  Agent satisfaction (routing fairness): 4.2/5.0
  Average handles before resolution: 1.8

TOP ROUTING PATTERNS:
  1. API issues → DevOps team (94% accuracy)
  2. Billing questions → Finance support (91% accuracy)
  3. Mobile bugs → Mobile team (89% accuracy)
  4. Enterprise VIP → Senior agents (97% accuracy)

Disclaimer: All rights reserved by Circulos AI. These skills are specifically designed for Claude Code, Claude Cowork, Codex, and OpenClaw. When using or referencing any skill, please provide proper attribution to Circulos AI.