<?xml version="1.0" encoding="UTF-8"?>
<rss xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:media="http://search.yahoo.com/mrss/" version="2.0"><channel><title>柒亦的小收获</title><link>http://www.qiyi.click</link><atom:link href="http://www.qiyi.click/rss.xml" rel="self" type="application/rss+xml"/><description>柒亦的小收获</description><generator>Halo v2.23.2</generator><language>zh-cn</language><lastBuildDate>Sat, 11 Apr 2026 05:00:25 GMT</lastBuildDate><item><title><![CDATA[把 AI 客服系统真正跑起来：aibotchat 的部署与监控实践]]></title><link>http://www.qiyi.click/archives/aibotchat-docker-deployment-and-observability</link><description><![CDATA[<img src="http://www.qiyi.click/plugins/feed/assets/telemetry.gif?title=%E6%8A%8A%20AI%20%E5%AE%A2%E6%9C%8D%E7%B3%BB%E7%BB%9F%E7%9C%9F%E6%AD%A3%E8%B7%91%E8%B5%B7%E6%9D%A5%EF%BC%9Aaibotchat%20%E7%9A%84%E9%83%A8%E7%BD%B2%E4%B8%8E%E7%9B%91%E6%8E%A7%E5%AE%9E%E8%B7%B5&amp;url=/archives/aibotchat-docker-deployment-and-observability" width="1" height="1" alt="" style="opacity:0;">这个项目让我重新认识了一件事：AI 系统能不能交付，最后看的不是模型多聪明，而是服务能不能稳定跑、出了问题能不能及时看见。]]></description><guid isPermaLink="false">/archives/aibotchat-docker-deployment-and-observability</guid><dc:creator>柒亦</dc:creator><category>AI客服中台</category><pubDate>Sat, 11 Apr 2026 00:53:58 GMT</pubDate></item><item><title><![CDATA[客服知识库检索实战：BM25、向量检索、RRF 与 Reranker 的组合]]></title><link>http://www.qiyi.click/archives/customer-service-rag-bm25-vector-rrf-reranker</link><description><![CDATA[<img src="http://www.qiyi.click/plugins/feed/assets/telemetry.gif?title=%E5%AE%A2%E6%9C%8D%E7%9F%A5%E8%AF%86%E5%BA%93%E6%A3%80%E7%B4%A2%E5%AE%9E%E6%88%98%EF%BC%9ABM25%E3%80%81%E5%90%91%E9%87%8F%E6%A3%80%E7%B4%A2%E3%80%81RRF%20%E4%B8%8E%20Reranker%20%E7%9A%84%E7%BB%84%E5%90%88&amp;url=/archives/customer-service-rag-bm25-vector-rrf-reranker" width="1" height="1" alt="" style="opacity:0;">我做这个项目时踩过一个很典型的坑：总以为接上向量检索，RAG 就差不多了。后来我才发现，客服知识库远没这么简单。]]></description><guid isPermaLink="false">/archives/customer-service-rag-bm25-vector-rrf-reranker</guid><dc:creator>柒亦</dc:creator><category>AI客服中台</category><pubDate>Sat, 11 Apr 2026 00:53:54 GMT</pubDate></item><item><title><![CDATA[从 FastAPI 到 RAG：我如何搭建一个 AI 客服中台]]></title><link>http://www.qiyi.click/archives/fastapi-rag-ai-customer-service-platform</link><description><![CDATA[<img src="http://www.qiyi.click/plugins/feed/assets/telemetry.gif?title=%E4%BB%8E%20FastAPI%20%E5%88%B0%20RAG%EF%BC%9A%E6%88%91%E5%A6%82%E4%BD%95%E6%90%AD%E5%BB%BA%E4%B8%80%E4%B8%AA%20AI%20%E5%AE%A2%E6%9C%8D%E4%B8%AD%E5%8F%B0&amp;url=/archives/fastapi-rag-ai-customer-service-platform" width="1" height="1" alt="" style="opacity:0;">这篇文章想写的不是技术名词，而是我在做 aibotchat 时，为什么决定把多轮对话、RAG、工具调用和转人工一起收进同一个客服系统。]]></description><guid isPermaLink="false">/archives/fastapi-rag-ai-customer-service-platform</guid><dc:creator>柒亦</dc:creator><category>AI客服中台</category><pubDate>Sat, 11 Apr 2026 00:53:49 GMT</pubDate></item></channel></rss>