AWS Labs▌

about

The Multi-Agent Orchestrator framework is a flexible and powerful tool for managing multiple AI agents and handling complex conversations. It's implemented in both Python and TypeScript, offering intelligent intent classification to route queries to the most appropriate agent. It supports both streaming and non-streaming responses, manages conversation context across agents, and boasts an extensible architecture for easy integration of new agents or customization of existing ones. Deployment is universal, working from AWS Lambda to local environments or any cloud platform.

features & capabilities

• /The Multi-Agent Orchestrator framework intelligently routes user queries to the most appropriate agents, maintaining contextual awareness throughout interactions.
• /The framework classifies user requests using an LLM, analyzing user requests, agent descriptions, and conversation history to understand ongoing conversations and context across all agents.
• /The orchestrator automatically handles saving the user’s input and the agent’s response into the storage for that specific user ID and session ID.
• /The framework provides several built-in agents for common tasks, allowing customization of properties and creation of custom agents for specific needs.
• /The orchestrator supports flexible storage options (in-memory and DynamoDB) and allows for custom storage solutions.
• /The framework includes built-in retrievers to enhance LLM-based agents’ performance by providing context and relevant information.

industry focus