EDDI

•
393
•

Orchestrates multi-agent conversations and business logic, providing intelligent routing, conversation management, and API orchestration for sophisticated AI-powered applications.

EDDI (Enhanced Dialog Driven Interface) is a high-performance multi-agent orchestration middleware designed to coordinate complex interactions between users, AI agents (LLMs), and business systems. Unlike standalone chatbots, EDDI acts as an intelligent layer, offering sophisticated conversation management, configurable behavior rules, and robust API orchestration for building advanced AI-powered applications. Developed in Java using Quarkus, it is lean, RESTful, scalable, and cloud-native, easily deployable with Docker and Kubernetes, and manages conversations through a unique Lifecycle Pipeline architecture with composable, version-controlled JSON configurations.

主要功能

01Stateful Conversation Management: Maintains complete, context-aware conversation history across multiple agents using a dedicated conversation memory.
02Configurable Behavior Rules: Defines when and how agents are invoked and business logic is executed through flexible IF-THEN rules, rather than hard-coded logic.
03Secure API Orchestration: Exposes pre-configured HTTP calls as secure tools for AI agents, providing a sandboxed and auditable layer for integrating with external APIs.
04288 GitHub stars
05Multi-Agent & LLM Orchestration: Intelligently routes conversations to various LLMs (OpenAI, Claude, Gemini, Ollama, Jlama) and orchestrates their invocation based on context and rules.
06Built-in AI Agent Tooling: Provides a suite of secure, autonomously invokable tools for AI agents, including calculator, web search, weather, and data formatting.

使用案例

01Developing intelligent virtual assistants or customer service bots requiring complex, multi-turn conversations and backend logic coordination.
02Building sophisticated AI-powered applications that integrate multiple LLMs with existing business systems and APIs.
03Managing conversations across various AI agents while maintaining state, context, and applying business-specific routing rules.