Skip to content

RasaHQ/agentic-orchestration-samples

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

35 Commits
 
 
 
 
 
 
 
 

Repository files navigation

Agentic Orchestration Samples

This repository contains runnable examples demonstrating Rasa's advanced agentic orchestration capabilities. These samples showcase how to build sophisticated conversational AI systems that can coordinate multiple agents, integrate external tools, and manage complex workflows through intelligent orchestration.

What is Agentic Orchestration?

Agentic orchestration refers to the ability to coordinate multiple specialized AI agents within a single conversational interface. This includes:

  • Multi-agent coordination: Managing conversations across different specialized agents
  • Protocol integration: Combining MCP (Model Context Protocol), A2A (Agent-to-Agent), and Rasa Flows
  • Context management: Maintaining conversation state and context across different phases
  • Workflow orchestration: Intelligent routing and decision-making between different conversation phases

Available Samples

1. Appointment Booking Assistant

A demonstration of how to use an agent to intelligently fill appointment slots based on user preferences and constraints. This sample showcases:

  • ReAct style sub agents in a flow: Using an MCP server to build a specialized reAct style sub agent for letting the user freely give their preferences and letting them pick an available slot.
  • Direct Tool Integration: Calling tools from an MCP server directly from Flows

📖 View README

2. Car Purchase Assistant

A demonstration of advanced agentic orchestration for complex, multi-phase workflows. This sample showcases:

  • MCP-Powered Research: Uses an agent that leverages an MCP server to perform car research (either real-time or using a mock dataset) and help users find suitable car models.
  • A2A-Powered Shopping: Integrates an A2A (Agent-to-Agent) server to power a dedicated car shopping agent, enabling structured car searches, recommendations, and dealer connections.
  • End-to-End Workflow: Guides users through research, shopping, and financing phases within a single conversational interface.
  • Context Management: Maintains and transfers relevant information seamlessly across all phases of the car buying journey.

📖 View README

Getting Started

Choose a sample that best fits your use case and follow its README for setup and usage instructions.

About

A collection of runnable sample assistants demonstrating Rasa’s agentic orchestration.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Contributors 6