🧩 mall-ai-after-sales-platform: AI for E-commerce After-Sales

The direction here is clear: building a trusted AI processing platform specifically for e-commerce after-sales scenarios. Instead of chasing generalized AI capabilities, it focuses on safely deploying AI in enterprise environments by combining RAG, controlled tool execution, and human-in-the-loop confirmation. The interesting part is its architectural split. It uses Java (Spring Boot) as the authoritative source for business logic and data writes, while leveraging LangGraph and MCP (Model Context Protocol) to orchestrate the AI workflows. This reflects a common requirement when integrating AI into traditional backend systems: balancing the flexibility of LLMs with the strict safety of database operations. It serves as a business-driven reference implementation. For developers figuring out how to introduce LLMs into existing Java applications without risking data corruption from AI hallucinations, this project offers a pragmatic approach featuring observability and human oversight.