研的轩轩的案例列表
LangGraph 多智能体客服
Production-oriented LangGraph customer-service agent with hybrid RAG, durable async execution, HITL governance, MCP tools, observability, and reproducible evaluation.一个独立维护的 LangGraph 多智能体客服工程参考实现：四层状态机动态路由 （缓存检查 → 意图路由 → 专家 Agent 协作 → 响应后处理），叠加分布式执行底座 （PostgreSQL checkpoint + Redis 锁 + Celery worke
行业 RAG 问答系统
Multimodal RAG QA for cosmetics regulation: hybrid retrieval, two-stage reranking, evidence gating that refuses unsupported answers, uint32 RBAC, evaluation and observability. FastAPI + Qdrant/Elasticsearch/Redis; evidence levels separate implemented from production-validated.项目边界：这是面向化妆品知识问答场景的公开、独
案例判断参考
研的轩轩的案例列表
LangGraph 多智能体客服
Production-oriented LangGraph customer-service agent with hybrid RAG, durable async execution, HITL governance, MCP tools, observability, and reproducible evaluation.一个独立维护的 LangGraph 多智能体客服工程参考实现：四层状态机动态路由 （缓存检查 → 意图路由 → 专家 Agent 协作 → 响应后处理），叠加分布式执行底座 （PostgreSQL checkpoint + Redis 锁 + Celery worke...
行业 RAG 问答系统
Multimodal RAG QA for cosmetics regulation: hybrid retrieval, two-stage reranking, evidence gating that refuses unsupported answers, uint32 RBAC, evaluation and observability. FastAPI + Qdrant/Elasticsearch/Redis; evidence levels separate implemented from production-validated.项目边界：这是面向化妆品知识问答场景的公开、独...
