LLMOps AI Application Platform
A Dify-like platform for orchestration, RAG, agents, and visual workflows
Python Flask Vue 3 LangGraph Weaviate Celery · May–Aug. 2026
An end-to-end LLMOps platform covering AI application orchestration, RAG, agents, and visual workflows behind a unified multi-model interface.
Function Calling and ReAct agents are built on LangGraph. The RAG pipeline is asynchronous — document parsing, chunking, Celery-driven indexing, and Weaviate hybrid retrieval, with long-term memory and multi-turn tool use. A visual DAG workflow engine handles dependency validation and execution debugging, and a unified adapter layer fronts OpenAI, DeepSeek, and Qwen.