jeecgboot/jeecg-ai

# Technical Summary Built on LangChain4j and Spring Boot 3.5.5, this platform combines RAG-powered AI workflows with a visual flow editor supporting multi-model backends (OpenAI, DeepSeek, Qwen, Claude, local Ollama) and vector database integration via pgvector. The backend uses Java/MyBatisPlus while the frontend leverages Vue3+TypeScript+Vite, enabling developers to compose reusable AI service APIs through canvas-based workflow orchestration with specialized nodes (AI inference, knowledge retrieval, branching logic, HTTP requests, custom Java/Python scripts). Native support for markdown document import with image preservation, PDF/PPT text extraction, and iframe-based chat embedding differentiates it from similar platforms, while integrated zero-code form builders and JimuReport analytics enable rapid business application assembly beyond pure AI interactions.