Agentic Design

Patterns
📋

Rule-based Routing(RBR)

A deterministic routing system that uses predefined rules, conditions, and decision trees to direct queries to appropriate handlers, providing fast, predictable, and auditable routing decisions through if-else statements, switch cases, and pattern matching ideal for compliance-critical and latency-sensitive agentic AI systems

Complexity: lowRouting

🎯 30-Second Overview

Pattern: Deterministic routing using predefined rules, conditions, and decision trees

Why: Provides traceable, auditable, ultra-fast routing decisions with guaranteed outcomes

Key Insight: If-else chains + pattern matching + config management = predictable agent orchestration

⚡ Quick Implementation

1Define:Create routing rules and conditions
2Extract:Parse keywords/patterns from input
3Match:Apply if-else/switch logic
4Validate:Check rule precedence and conflicts
5Execute:Route to deterministic handler
Example: if (query.contains("refund")) route("billing_agent") else if (priority == "urgent") route("escalation_handler")

📋 Do's & Don'ts

✅Order rules from most specific to general
✅Implement fallback/default routes
✅Version control your rule definitions
✅Log routing decisions for debugging
✅Combine with AI for hybrid workflows
❌Create overlapping or conflicting rules
❌Hardcode rules without config management
❌Ignore edge cases and exceptions
❌Over-complicate with too many conditions
❌Skip validation and testing of rules

🚦 When to Use

Use When

  • • Deterministic routing is required
  • • Compliance/audit trails needed
  • • Low-latency decisions critical
  • • Clear, stable routing patterns

Avoid When

  • • Complex semantic understanding needed
  • • Patterns frequently change
  • • Handling ambiguous queries
  • • Natural language nuance matters

📊 Key Metrics

Rule Coverage
% queries matched by rules
Decision Latency
Time to route (usually <5ms)
Rule Conflicts
Number of overlapping rules
Fallback Rate
% routed to default handler
Rule Maintenance
Changes per month
Accuracy
% correctly routed vs manual

💡 Top Use Cases

Priority Escalation: if urgency="critical" AND type="security" → security_team
Department Routing: switch(category) { case "billing": finance_agent; case "tech": it_support }
Compliance Workflows: regulated industries requiring deterministic, auditable paths
Load Distribution: round-robin or weighted routing based on agent availability
Hybrid Orchestration: rules for known patterns + AI fallback for complex cases

References & Further Reading

Deepen your understanding with these curated resources

Academic Papers & Research

AI Agent Orchestration Patterns - Microsoft Azure Architecture

Deterministic vs Dynamic Process Orchestration in AI (2024)

AI Agents and Deterministic Workflows: A Spectrum - deepset

Agentic AI Design Patterns & Real-World Examples (2024)

Implementation Guides

Ultimate Guide to AI Agent Routing - Botpress

Workflows & Agents Tutorial - LangGraph

Agentic Workflows Architecture - Orkes

Building Scalable AI: Workflows vs Agents

Tools & Frameworks

Apache Airflow - Workflow Orchestration

Temporal - Durable Workflow Execution

Camunda - Process Orchestration Platform

n8n - Workflow Automation Tool

Industry Reports & Best Practices

Seizing the Agentic AI Advantage - McKinsey

Rise of Agentic AI - Capgemini Report

Agentic Workflows Patterns - Weaviate

9 Agentic AI Workflow Patterns (2024) - MarkTechPost

Contribute to this collection

Know a great resource? Submit a pull request to add it.

Contribute

Patterns

closed

Loading...