<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
<channel>
  <title>AI 武林：LangSmith</title>
  <link>https://aiwulin.itsmygo.uk/tools/langsmith/</link>
  <description>AI 武林收錄的內容裡，最新提到 LangSmith 的 30 筆，每筆附中文摘要。</description>
  <language>zh-TW</language>
  <lastBuildDate>Thu, 08 Oct 2026 17:22:41 GMT</lastBuildDate>
  <atom:link href="https://aiwulin.itsmygo.uk/tools/langsmith/rss.xml" rel="self" type="application/rss+xml"/>
  <item>
    <title>Building Engine v2: How LangChain approaches agent improvement</title>
    <link>https://aiwulin.itsmygo.uk/c/03ae7945e8/</link>
    <guid isPermaLink="false">aiwulin-03ae7945e8</guid>
    <pubDate>Wed, 07 Oct 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;Ben Tannyhill 介紹 LangSmith Engine 如何透過自我評估與紅隊測試提升代理效能，並分享跨追蹤趨勢檢測與小模型優勢等實作經驗。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=S3l7U7gOsyY&quot;&gt;https://www.youtube.com/watch?v=S3l7U7gOsyY&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>開發框架</category>
    <category>AI Agent 基礎</category>
    <category>AI 評測</category>
  </item>
  <item>
    <title>How to Build a Model Router in the Harness</title>
    <link>https://aiwulin.itsmygo.uk/c/09019aca8c/</link>
    <guid isPermaLink="false">aiwulin-09019aca8c</guid>
    <pubDate>Wed, 07 Oct 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangChain 產品經理 Sydney Runkle 示範如何將模型路由整合到 Open SWE 程式碼代理中，透過分析任務分佈與選擇帕累託前沿模型來降低成本。看完能學會如何建立路由機制並透過 A/B 測試驗證效果。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=5kTFyEOgark&quot;&gt;https://www.youtube.com/watch?v=5kTFyEOgark&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>開發框架</category>
    <category>Agent Skills 與 Harness</category>
    <category>開源模型</category>
  </item>
  <item>
    <title>How Zip Uses LangSmith and LangGraph to Ship New Features Faster</title>
    <link>https://aiwulin.itsmygo.uk/c/3d126cb7fd/</link>
    <guid isPermaLink="false">aiwulin-3d126cb7fd</guid>
    <pubDate>Tue, 06 Oct 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;Zip 團隊分享如何透過 LangGraph 與 LangSmith 加速 AI 功能開發與追蹤。透過這些工具，團隊無需額外編碼即可獲得節點互動與 LLM 呼叫的追蹤資料，並建立評估機制防止回退問題。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=sGrD9a8dh_0&quot;&gt;https://www.youtube.com/watch?v=sGrD9a8dh_0&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>開發框架</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>Add Web Search to Managed Deep Agents for free with One MCP Config</title>
    <link>https://aiwulin.itsmygo.uk/c/4765edf4e6/</link>
    <guid isPermaLink="false">aiwulin-4765edf4e6</guid>
    <pubDate>Tue, 06 Oct 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangChain 產品經理演示如何透過單一 MCP 設定，免費為管理式深度代理加入網頁搜尋功能。影片展示從內部文件受限到整合搜尋伺服器，並在 LangSmith 中部署與檢視工具呼叫的完整流程。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=KAITWbHLSSE&quot;&gt;https://www.youtube.com/watch?v=KAITWbHLSSE&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>MCP 模型上下文協定</category>
    <category>AI Agent 基礎</category>
    <category>開發框架</category>
  </item>
  <item>
    <title>Interrupt NYC: Opening Keynote</title>
    <link>https://aiwulin.itsmygo.uk/c/6f7fa0b31d/</link>
    <guid isPermaLink="false">aiwulin-6f7fa0b31d</guid>
    <pubDate>Mon, 05 Oct 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;Harrison Chase 分享如何透過 LangGraph 與 LangSmith 建構領域專屬智慧，並介紹 Managed Deep Agents、LangSmith Engine v2 等最新功能與觀察工具。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=950byF7njfw&quot;&gt;https://www.youtube.com/watch?v=950byF7njfw&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>開發框架</category>
  </item>
  <item>
    <title>HTTP Channels for Managed Deep Agents</title>
    <link>https://aiwulin.itsmygo.uk/c/d980a0dbd8/</link>
    <guid isPermaLink="false">aiwulin-d980a0dbd8</guid>
    <pubDate>Thu, 01 Oct 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangChain 產品經理 Nathan Drezner 示範如何為管理型深度代理建立自訂 HTTP 通道，讓代理能接收來自外部服務（如 HubSpot、Salesforce）的請求。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=6v0Fmsi4Ldk&quot;&gt;https://www.youtube.com/watch?v=6v0Fmsi4Ldk&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
    <category>開發框架</category>
  </item>
  <item>
    <title>Evaluating Agents in Production: Traces, LLM-as-Judge, and Prompt Management at Wonder</title>
    <link>https://aiwulin.itsmygo.uk/c/09c791cb31/</link>
    <guid isPermaLink="false">aiwulin-09c791cb31</guid>
    <pubDate>Tue, 29 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;Wonder 團隊分享如何將人工除錯改為使用 LangSmith 追蹤與 LLM 評分，並透過自動化回饋迴圈最佳化提示詞管理。觀眾可學習如何建立可擴展的 AI 應用評估與維護流程。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=kRaJtEXAOxU&quot;&gt;https://www.youtube.com/watch?v=kRaJtEXAOxU&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI 評測</category>
    <category>大型語言模型入門</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>How to go from your agent's traces to a fine-tuned model in one workflow</title>
    <link>https://aiwulin.itsmygo.uk/c/3a2d00527b/</link>
    <guid isPermaLink="false">aiwulin-3a2d00527b</guid>
    <pubDate>Thu, 24 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;介紹如何使用 smithtune 工具將 LangSmith 的追蹤記錄轉化為訓練資料，並對 Qwen 模型進行監督式微調。觀眾可學習如何建立金標資料集、評估軌跡品質，並完成從資料準備到模型部署的完整工作流。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=pb-AAvvQZ-U&quot;&gt;https://www.youtube.com/watch?v=pb-AAvvQZ-U&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>微調</category>
    <category>可觀測性與 LLMOps</category>
    <category>AI Agent 基礎</category>
  </item>
  <item>
    <title>Introducing LangSmith Trajectories: A readable view of every agent session</title>
    <link>https://aiwulin.itsmygo.uk/c/f33026000d/</link>
    <guid isPermaLink="false">aiwulin-f33026000d</guid>
    <pubDate>Thu, 24 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangSmith 推出 Trajectories，以時間軸方式呈現 AI 代理的完整對話路徑，讓使用者能快速定位問題根源並進行評估。此功能協助專家以易讀格式審查代理行為，並支援將資料用於監督微調。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=1akSvsjd_oI&quot;&gt;https://www.youtube.com/watch?v=1akSvsjd_oI&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>Introducing LangSmith Custom Apps: Build custom interfaces around your agent data</title>
    <link>https://aiwulin.itsmygo.uk/c/69d60fe550/</link>
    <guid isPermaLink="false">aiwulin-69d60fe550</guid>
    <pubDate>Thu, 24 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangSmith 推出自訂應用程式功能，讓使用者能基於代理資料建立、發布並執行自訂介面。透過模板、程式碼代理指導或提示即可快速構建，無需從零開始。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=aMqBsTRD2XA&quot;&gt;https://www.youtube.com/watch?v=aMqBsTRD2XA&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>Schedules for Managed Deep Agents: Cron jobs, prompts, and Slack delivery</title>
    <link>https://aiwulin.itsmygo.uk/c/5ac26eecd2/</link>
    <guid isPermaLink="false">aiwulin-5ac26eecd2</guid>
    <pubDate>Thu, 24 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangChain 產品團隊示範如何為管理型深度代理設定排程，使其能透過 cron 任務自動執行並將結果傳送至 Slack。觀眾可學習如何建立週期性任務，讓 AI 代理在無需人工干預的情況下定期運作。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=-mWWsvv19jE&quot;&gt;https://www.youtube.com/watch?v=-mWWsvv19jE&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
    <category>提示工程</category>
  </item>
  <item>
    <title>Introducing LangSmith Engine v2: Detect and ship fixes to agents faster</title>
    <link>https://aiwulin.itsmygo.uk/c/db03a364b2/</link>
    <guid isPermaLink="false">aiwulin-db03a364b2</guid>
    <pubDate>Thu, 24 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangSmith 推出 Engine v2，強化自動偵測與驗證 AI 代理問題的能力。新功能包含主動紅隊測試、識別更多難發現問題型別，以及自動測試修復方案並快速部署。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=GNO9XFE4PKM&quot;&gt;https://www.youtube.com/watch?v=GNO9XFE4PKM&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
    <category>可觀測性與 LLMOps</category>
    <category>AI 資安</category>
  </item>
  <item>
    <title>LangSmith explained in 5 minutes</title>
    <link>https://aiwulin.itsmygo.uk/c/512bdea7f0/</link>
    <guid isPermaLink="false">aiwulin-512bdea7f0</guid>
    <pubDate>Wed, 23 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;介紹 LangSmith 平台，涵蓋從建立、測試到監控代理的全流程。觀眾可學習如何追蹤執行、分析錯誤並部署 AI 系統。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=KPcEhLoh-Wc&quot;&gt;https://www.youtube.com/watch?v=KPcEhLoh-Wc&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>Ship a GitHub PR From a Slack Message with Managed Deep Agents</title>
    <link>https://aiwulin.itsmygo.uk/c/4d2ed92346/</link>
    <guid isPermaLink="false">aiwulin-4d2ed92346</guid>
    <pubDate>Wed, 23 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangChain 團隊展示 Patch，一個將 Slack 訊息轉為 GitHub PR 的代理工具，並演示如何用 Managed Deep Agents 快速開發 Tetris 側專案功能。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=4-oA1BJeyr4&quot;&gt;https://www.youtube.com/watch?v=4-oA1BJeyr4&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
    <category>開發框架</category>
    <category>Coding Agent</category>
  </item>
  <item>
    <title>Building a Harness with Jev</title>
    <link>https://aiwulin.itsmygo.uk/c/c7bb5f3b26/</link>
    <guid isPermaLink="false">aiwulin-c7bb5f3b26</guid>
    <pubDate>Tue, 22 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;介紹 TypeSafe AI 推出的 Jev 模型，一種比傳統大型語言模型快且便宜的 System 1 決策型 AI。影片示範如何用 Jev 進行模型路由、自動過濾風險工具呼叫，以及作為線上評估的評分者，並提供在 LangChain 中的…&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=VE5dsWll06M&quot;&gt;https://www.youtube.com/watch?v=VE5dsWll06M&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI 評測</category>
    <category>Agent Skills 與 Harness</category>
    <category>開發框架</category>
  </item>
  <item>
    <title>Middleware for Managed Deep Agents</title>
    <link>https://aiwulin.itsmygo.uk/c/ea03aae3f3/</link>
    <guid isPermaLink="false">aiwulin-ea03aae3f3</guid>
    <pubDate>Thu, 17 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangChain 產品經理 Nathan Drezner 介紹管理式深度代理的中介層，示範如何加入預設的 PII 中介層來遮蔽客戶郵件，並從零編寫自訂的日誌中介層。觀眾能學習如何為代理擴展生命週期，加入政策執行、容錯與速率限制等自訂行為。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=d6aNbE-3dxo&quot;&gt;https://www.youtube.com/watch?v=d6aNbE-3dxo&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
    <category>開發框架</category>
  </item>
  <item>
    <title>How Lyft Increased Its Agent Resolution Rate by 16% with LangSmith and LangGraph</title>
    <link>https://aiwulin.itsmygo.uk/c/3876f1f3dd/</link>
    <guid isPermaLink="false">aiwulin-3876f1f3dd</guid>
    <pubDate>Wed, 16 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;Lyft 工程師分享如何利用 LangGraph 與 LangSmith 重建客服代理架構，將開發時間從六個月縮短至兩週，並將解決率提升 16%。透過元代理架構與可觀測性工具，讓產品經理與營運人員無需寫程式即可快速部署新代理。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=M9BMTC8o9-w&quot;&gt;https://www.youtube.com/watch?v=M9BMTC8o9-w&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>開發框架</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>Share your Managed Deep Agent with your team using Slack</title>
    <link>https://aiwulin.itsmygo.uk/c/c15cad5a83/</link>
    <guid isPermaLink="false">aiwulin-c15cad5a83</guid>
    <pubDate>Tue, 15 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;示範如何將 LangSmith 中的研究助理代理部署到 Slack 團隊環境，包含設定連線、授權、監控追蹤及自訂外觀。看完即可掌握將 AI 代理整合至 Slack 並讓團隊使用的具體步驟。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=a5Yie-Bgx7A&quot;&gt;https://www.youtube.com/watch?v=a5Yie-Bgx7A&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>Catch Agent Regressions Before You Ship: Evals for Managed Deep Agents</title>
    <link>https://aiwulin.itsmygo.uk/c/0812a387b1/</link>
    <guid isPermaLink="false">aiwulin-0812a387b1</guid>
    <pubDate>Fri, 11 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangChain 產品團隊成員 Nathan 介紹 Managed Deep Agents 內建的評估能力，說明如何透過 Harbor 環境、任務與檢查機制，自動檢測模型退化並監控執行狀態。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=pjv-mp5J4hA&quot;&gt;https://www.youtube.com/watch?v=pjv-mp5J4hA&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI 評測</category>
    <category>AI Agent 基礎</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>Build an Online LLM-as-a-Judge with the LangSmith CLI and a Coding Agent</title>
    <link>https://aiwulin.itsmygo.uk/c/44ae645270/</link>
    <guid isPermaLink="false">aiwulin-44ae645270</guid>
    <pubDate>Fri, 11 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;示範如何使用 LangSmith CLI 與編碼代理建立線上 LLM 評審系統。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=XE5-hb-H-OY&quot;&gt;https://www.youtube.com/watch?v=XE5-hb-H-OY&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>Coding Agent</category>
    <category>可觀測性與 LLMOps</category>
    <category>大型語言模型入門</category>
  </item>
  <item>
    <title>Inside Clay's Eval Stack: 300M Agent Runs, One LangSmith Pipeline</title>
    <link>https://aiwulin.itsmygo.uk/c/e417911686/</link>
    <guid isPermaLink="false">aiwulin-e417911686</guid>
    <pubDate>Wed, 02 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;介紹 Clay 的評估流程，包含 3 億次代理執行與單一 LangSmith 管道。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=Uny6LpmjraI&quot;&gt;https://www.youtube.com/watch?v=Uny6LpmjraI&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI 評測</category>
    <category>可觀測性與 LLMOps</category>
    <category>AI Agent 基礎</category>
  </item>
  <item>
    <title>Accelerate agent improvement with LangSmith Engine</title>
    <link>https://aiwulin.itsmygo.uk/c/8c804373e0/</link>
    <guid isPermaLink="false">aiwulin-8c804373e0</guid>
    <pubDate>Wed, 02 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;介紹如何使用 LangSmith Engine 加速 Agent 的改進。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=hYhTke2m0X8&quot;&gt;https://www.youtube.com/watch?v=hYhTke2m0X8&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>Introducing: LangSmith Tuned Evaluators</title>
    <link>https://aiwulin.itsmygo.uk/c/bc24ba6f5d/</link>
    <guid isPermaLink="false">aiwulin-bc24ba6f5d</guid>
    <pubDate>Wed, 02 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;介紹 LangSmith 的自動調整評估器功能。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=0oh0X2NQgQo&quot;&gt;https://www.youtube.com/watch?v=0oh0X2NQgQo&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI 評測</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>LangSmith Preview Builds: Test agent changes before production</title>
    <link>https://aiwulin.itsmygo.uk/c/f2ec400087/</link>
    <guid isPermaLink="false">aiwulin-f2ec400087</guid>
    <pubDate>Wed, 02 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;LangSmith 預覽版本更新，讓使用者能在生產環境前測試 Agent 變更。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=xEnUDGZ3_HE&quot;&gt;https://www.youtube.com/watch?v=xEnUDGZ3_HE&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>可觀測性與 LLMOps</category>
    <category>AI Agent 基礎</category>
  </item>
  <item>
    <title>Keep model costs in check with LangSmith LLM Gateway</title>
    <link>https://aiwulin.itsmygo.uk/c/770446c76a/</link>
    <guid isPermaLink="false">aiwulin-770446c76a</guid>
    <pubDate>Wed, 02 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;介紹如何用 LangSmith LLM Gateway 控制模型成本。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=VBZgp1sHosE&quot;&gt;https://www.youtube.com/watch?v=VBZgp1sHosE&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>大型語言模型入門</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>Inside Toyota's Manufacturing AI: LangChain, LangSmith, and Six Figure ROI</title>
    <link>https://aiwulin.itsmygo.uk/c/9a9b81e16c/</link>
    <guid isPermaLink="false">aiwulin-9a9b81e16c</guid>
    <pubDate>Wed, 02 Sep 2026 00:00:00 +0800</pubDate>
    <dc:creator>LangChain</dc:creator>
    <description>&lt;p&gt;介紹 Toyota 如何利用 LangChain 與 LangSmith 打造製造業 AI，實現六位數投資報酬率。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=Z9FenPGLoLw&quot;&gt;https://www.youtube.com/watch?v=Z9FenPGLoLw&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>開發框架</category>
    <category>可觀測性與 LLMOps</category>
  </item>
  <item>
    <title>Voice Agent observability with LangSmith</title>
    <link>https://aiwulin.itsmygo.uk/c/b4dc090606/</link>
    <guid isPermaLink="false">aiwulin-b4dc090606</guid>
    <pubDate>Sun, 02 Aug 2026 00:00:00 +0800</pubDate>
    <dc:creator>Google for Developers</dc:creator>
    <description>&lt;p&gt;介紹如何使用 LangSmith 監控語音代理的效能與問題。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=z6l5nyE_6-I&quot;&gt;https://www.youtube.com/watch?v=z6l5nyE_6-I&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>可觀測性與 LLMOps</category>
    <category>語音與音訊</category>
  </item>
  <item>
    <title>Agentic AI – Complete Course for Beginners</title>
    <link>https://aiwulin.itsmygo.uk/c/eeda2c0cc9/</link>
    <guid isPermaLink="false">aiwulin-eeda2c0cc9</guid>
    <pubDate>Thu, 30 Jul 2026 00:00:00 +0800</pubDate>
    <dc:creator>freeCodeCamp.org</dc:creator>
    <description>&lt;p&gt;教導如何從零開始構建生產級多代理系統，掌握 LangChain 與 LangGraph 的核心概念與實作技巧。學員將學習如何設計序列、並行與條件工作流，整合 RAG、記憶機制與人工介入控制，並透過實際專案將應用程式部署至 AWS 與…&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://www.youtube.com/watch?v=Zy7EXDONlTY&quot;&gt;https://www.youtube.com/watch?v=Zy7EXDONlTY&lt;/a&gt;&lt;/p&gt;</description>
    <category>影片</category>
    <category>AI Agent 基礎</category>
    <category>開發框架</category>
    <category>多代理系統</category>
  </item>
  <item>
    <title>Everything Gets Rebuilt: The New AI Agent Stack | Harrison Chase, LangChain</title>
    <link>https://aiwulin.itsmygo.uk/c/75b9327745/</link>
    <guid isPermaLink="false">aiwulin-75b9327745</guid>
    <pubDate>Thu, 12 Mar 2026 00:00:00 +0800</pubDate>
    <dc:creator>The MAD Podcast</dc:creator>
    <description>&lt;p&gt;Harrison Chase 與 LangChain 共同創辦人深入探討 AI 代理（Agent）如何從簡單提示演變為具備規劃、編碼與工具使用能力的複雜系統。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://podcasters.spotify.com/pod/show/firstmark/episodes/Everything-Gets-Rebuilt-The-New-AI-Agent-Stack--Harrison-Chase--LangChain-e3gapv9&quot;&gt;https://podcasters.spotify.com/pod/show/firstmark/episodes/Everything-Gets-Rebuilt-The-New-AI-Agent-Stack--Harrison-Chase--LangChain-e3gapv9&lt;/a&gt;&lt;/p&gt;</description>
    <category>Podcast</category>
    <category>AI Agent 基礎</category>
    <category>開發框架</category>
  </item>
  <item>
    <title>愛好 AI Engineer 電子報 🚀 新型態代理人 OpenClaw 正夯，電子報改版 #35</title>
    <link>https://aiwulin.itsmygo.uk/c/a40650c6c7/</link>
    <guid isPermaLink="false">aiwulin-a40650c6c7</guid>
    <pubDate>Sun, 22 Feb 2026 00:00:00 +0800</pubDate>
    <dc:creator>ihower（張文鈿）</dc:creator>
    <description>&lt;p&gt;介紹新型態開源 AI 代理軟體 OpenClaw 的用法與安全設定，並整理近期 AI Agent 架構、評估與效能最佳化等文章重點。讀者可學習如何部署個人助理、選擇多代理設計模式，以及系統化評估 AI 表現的方法。&lt;/p&gt;&lt;p&gt;原站：&lt;a href=&quot;https://ihower.tw/blog/13630-aie-openclaw&quot;&gt;https://ihower.tw/blog/13630-aie-openclaw&lt;/a&gt;&lt;/p&gt;</description>
    <category>文章</category>
    <category>AI Agent 基礎</category>
    <category>AI 輔助軟體工程</category>
  </item>
</channel>
</rss>
