Show HN: EDDI – Multi-agent AI engine where agent logic lives in JSON, not code

Category: ai-ml

Tags: multi-agent, conversational-ai, orchestration, java, quarkus, configuration-driven

Score: 7.5/10 (Innovation: 7, Technical: 8, Documentation: 8, Utility: 7)

EDDI is a production-grade, config-driven multi-agent orchestration middleware for conversational AI, built in Java/Quarkus. It's interesting because it inverts the typical Python/Node.js agent framework model by using deterministic JSON configurations for agent logic instead of embedded code, focusing on security, compliance, and governance for enterprise deployment.

Target audience: backend devs, devops, enterprise architects

Repository: https://github.com/labsai/EDDI · Java · Apache-2.0 · 309 stars

View on Hacker News