📄 中文摘要
构建大语言模型（LLM）应用程序不仅仅依赖于巧妙的提示，而是需要工程化的检索增强生成（RAG）管道。大多数教程仅展示了加载文档、嵌入、存储在向量数据库中以及向 GPT 提问的基本步骤。然而，在实际的 LLM 系统中，这仅仅是第一步。生产级的 RAG 需要多种复杂的技术，包括查询重写、分块策略、混合搜索、重排序、评估管道、安全防护、延迟优化和成本治理等。这些要素共同构成了真正的 LLM 工程，远超简单的提示使用。
教你真正的检索工程的 RAG 项目（不仅仅是提示黑客）
📄 English Summary
RAG Projects That Teach You Real Retrieval Engineering (Not Just Prompt Hacking)
Building applications with large language models (LLMs) is no longer just about clever prompts; it requires engineering robust Retrieval-Augmented Generation (RAG) pipelines. Most tutorials simplify the process to loading documents, embedding them, storing in a vector database, and asking GPT a question. However, in real-world LLM systems, this is merely the first step. Production-grade RAG necessitates various complex techniques, including query rewriting, chunking strategies, hybrid search, reranking, evaluation pipelines, guardrails, latency optimization, and cost governance. These elements collectively form the foundation of true LLM engineering, far beyond simple prompt usage.