MovieAgent Unveils Six-Layer Architecture and Vector Store Results
The MovieAgent project author published the second installment of a ten-part series, detailing a six-layer dependency model, a request path from HTTP input to streamed results, and three configuration switches for vector storage, embeddings, and LLM providers. The implementation uses a five-node LangGraph flow while reserving LLMs for intent parsing and response generation, with planning, execution, and fallback handling kept deterministic. On a self-reported sample of 1,134 films, Chroma, embedded Qdrant, and Milvus Lite showed sub-100-millisecond query latency, though the results have not been independently verified.
The blueprint provides concrete engineering guidance on isolating AI-agent components and switching infrastructure providers without changing business logic, while its vector-store measurements offer a practical comparison for local development.
第 2 篇 · 整体架构蓝图:六层依赖与技术选型
《从零造一个影片推荐智能体》系列 · 第 2 篇 / 共 10 篇 上篇撂下一句「没有最完美的架构,只有最贴合当前场景的取舍」,这篇就来还债——把那张地图摊开。
前言
第 1 篇结尾我说,这张地图具体怎么画、六层怎么摞、请求从 HTTP 进来怎么一路走到结构化结果,是下一篇的事。今儿就还这笔债。
写业务代码有个很容易上头的坑:一想清楚「要做个推荐 Agent」,手就痒,先 pip install,再顺手建几个包,边写边改目录。我见过太多项目(自己以前也没少干)最后是这么烂掉的——utils.py 里塞了数据库连接,api/ 反过来 import 了 agent/ 的内部函数,改一个召回权重得翻五个文件。等到想加个缓存,发现根本没地方下嘴。
所以这个项目的第一步,我硬是按住没写代码,先花了几天画一张图。今天真正要摊开的,是本篇的主角 -- 一张「谁依赖谁」的分层图,外加三个「把变化关进笼子」的开关。
#MovieAgent#LangGraph#LangChain#FastAPI#Chroma#Qdrant#Milvus Lite#Ollama#DeepSeek#bge-m3#Qwen#MySQL