Router

A central agent inspects each request and routes it to the most appropriate specialist agent based on content and context.

Overview

Routers keep front-line interactions lightweight: they triage requests, choose which specialist should respond, and optionally provide fallbacks when no match is found. Common use cases include customer support triage and language translation hubs.

Demo Scenario: Language Help Desk

Three agents collaborate:

• Spanish expert – handles Spanish translation and explanations

• Japanese expert – handles Japanese translation and explanations

• Router agent – analyzes each question and delegates to the correct expert

The router uses simple heuristics in its instruction to decide which expert to invoke.

Diagram

Implementation Steps

1. Create language experts

   from glaip_sdk import Agent
   
   spanish_expert = Agent(
       name="spanish_expert",
       instruction="You respond only in Spanish",
   )
   
   japanese_expert = Agent(
       name="japanese_expert",
       instruction="You are a Japanese language teacher. Provide friendly explanations and translations involving Japanese.",
   )

2. Create the router agent and point it at the specialists

   router_agent = Agent(
       name="router_agent",
       instruction="""Analyze the user's query and decide which expert to route to.
       Rules:
       1. Queries about Spanish → spanish_expert
       2. Queries about Japanese → japanese_expert
       3. Otherwise respond that no expert is available""",
       agents=[spanish_expert, japanese_expert],
   )

3. Process incoming requests

   queries = [
       "How do you say 'love' in Spanish?",
       "What is the meaning of 'arigatou' in English?",
       "How do you say 'hello' in German?",
   ]
   
   for query in queries:
       print(f"\n--- Processing Query: {query} ---")
       router_agent.run(query)

4. Clean up the demo agents

   # Only needed if agents were deployed
   # router_agent.delete()
   # spanish_expert.delete()
   # japanese_expert.delete()

How to Run

From the repository root glaip-sdk/python/glaip-sdk:

poetry run python examples/multi-agent-system-patterns/router/router.py

Set the usual environment variables in .env:

OPENAI_API_KEY=your-openai-key-here

Output

--- Processing Query: How do you say 'love' in Spanish? ---
• Routed to spanish_expert → "'Love' in Spanish is 'amor'. Here's how you use it..."

--- Processing Query: What is the meaning of 'arigatou' in English? ---
• Routed to japanese_expert → "'Arigatou' means 'thank you'. Here's the nuance..."

--- Processing Query: How do you say 'hello' in German? ---
• Router response → "No expert available for this query."

Notes

• Extend the instruction with additional heuristics (language detection, confidence scores) or add more specialists for other languages.

• Combine with an aggregator to produce a unified transcript of all routed interactions.

• Monitor routing outcomes and adjust the prompt rules as queries evolve.

Related Documentation

• Agents guide — Configure nested agents, memory, and streaming renderers.

• Automation & scripting — Log routing results or wire the router into scheduled jobs.

• Security & privacy — Apply memory and PII controls when routing sensitive requests.