<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Lena Nadir Blog</title><description>Welcome to my website!</description><link>https://www.lena-nadir.dev/</link><item><title>AI-103 Study Guide: How I Passed the Azure AI Apps &amp; Agents Exam on My First Try</title><link>https://www.lena-nadir.dev/blog/en/ai-103-exam-experience/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/en/ai-103-exam-experience/</guid><description>A first-attempt pass report for the Microsoft AI-103 (Azure AI Apps and Agents Developer Associate) exam — domain weighting, an LLM-assisted hands-on study method, and tips for exam day.</description><pubDate>Thu, 16 Jul 2026 00:00:00 GMT</pubDate></item><item><title>[AI-103] Generative AI/Agent Hands-On Roadmap</title><link>https://www.lena-nadir.dev/blog/en/ai-103-series-00-roadmap/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/en/ai-103-series-00-roadmap/</guid><description>The intro post for a hands-on series focused on the Generative AI/Agent domain of AI-103. Covers the overall theme and structure of the series.</description><pubDate>Sat, 18 Jul 2026 00:00:00 GMT</pubDate></item><item><title>[AI-103 Hands-On] Part 1: Building an Agent in Foundry Portal</title><link>https://www.lena-nadir.dev/blog/en/ai-103-series-01-agent-basics/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/en/ai-103-series-01-agent-basics/</guid><description>Build a no-code Agent with the Web Search Tool in Foundry Portal, then call that same Agent from the Python SDK (Responses API) — a full end-to-end walkthrough on real infrastructure.</description><pubDate>Tue, 21 Jul 2026 00:00:00 GMT</pubDate></item><item><title>[AI-103 Hands-On] Part 2: Implementing Basic RAG with File Search</title><link>https://www.lena-nadir.dev/blog/en/ai-103-series-02-file-search-rag/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/en/ai-103-series-02-file-search-rag/</guid><description>Upload documents with Microsoft Foundry&apos;s File Search feature and see an Agent answer using Retrieval-Augmented Generation (RAG) grounded in those files — confirmed hands-on via the Python SDK.</description><pubDate>Sat, 25 Jul 2026 00:00:00 GMT</pubDate></item><item><title>[AI-103 Hands-On] Part 3: How to Get Structured JSON Output</title><link>https://www.lena-nadir.dev/blog/en/ai-103-series-03-structured-output/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/en/ai-103-series-03-structured-output/</guid><description>We build two Foundry Agents with Web Search enabled — one Instructions-only, one Instructions plus a JSON Schema — so search and structuring happen in a single call, then compare them hands-on on parse success rate, type stability, and token cost.</description><pubDate>Sat, 01 Aug 2026 00:00:00 GMT</pubDate></item><item><title>What Is AI-500? Exam Scope, Differences from AI-103, and How It Compares with Other Certifications</title><link>https://www.lena-nadir.dev/blog/en/ai-500-beta-exam-trends/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/en/ai-500-beta-exam-trends/</guid><description>An in-depth look at Microsoft’s new AI-500 certification based on official sources, covering the exam scope, its relationship with AI-103, early reports on difficulty, LangGraph vs. Microsoft Agent Framework, and comparisons with AWS, Google Cloud, and NVIDIA certifications.</description><pubDate>Sun, 16 Aug 2026 00:00:00 GMT</pubDate></item><item><title>AI Agent Frameworks &amp; Platforms Compared: LangChain, Microsoft, AWS, and Google&apos;s Strategy and Strengths</title><link>https://www.lena-nadir.dev/blog/en/ai-agent-framework-comparison/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/en/ai-agent-framework-comparison/</guid><description>A comparison of LangChain/LangGraph, Microsoft Agent Framework, AWS (Bedrock Agents/Strands Agents), and Google&apos;s Agent Development Kit (ADK) across provider, strengths, state management approach, and LLM compatibility.</description><pubDate>Mon, 10 Aug 2026 00:00:00 GMT</pubDate></item><item><title>Data Platform Modernization and Solution Selection</title><link>https://www.lena-nadir.dev/blog/en/data-platform-modernization/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/en/data-platform-modernization/</guid><description>A practical account of why we revisited Azure Databricks after running a production-oriented PoC on Microsoft Fabric—and what the experience taught us about requirements, architectural portability, and solution selection for an AI-ready data platform.</description><pubDate>Sat, 05 Sep 2026 00:00:00 GMT</pubDate></item><item><title>Implementing the Generator &amp; Critic Pattern with LangGraph — An Agent Generates Unit Test Specifications</title><link>https://www.lena-nadir.dev/blog/en/langgraph-test-spec-generator/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/en/langgraph-test-spec-generator/</guid><description>Implement a Generator &amp; Critic workflow with LangGraph that supplements and reviews unit test specifications from design documents, source code, and developer-authored key test items.</description><pubDate>Sun, 30 Aug 2026 00:00:00 GMT</pubDate></item><item><title>Implementing the Parallel Fan-Out/Gather Pattern on Microsoft Foundry: A Stock Analysis Example</title><link>https://www.lena-nadir.dev/blog/en/multi-agent-stock-analysis/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/en/multi-agent-stock-analysis/</guid><description>Implementing a multi-agent Parallel Fan-Out/Gather Pattern on Microsoft Foundry. We organize the multi-agent design patterns from the Google Developers Blog, then use stock analysis as a use case to test running specialist agents in parallel and synthesizing their results.</description><pubDate>Wed, 26 Aug 2026 00:00:00 GMT</pubDate></item><item><title>AI-103合格体験記：一発合格までの学習方法</title><link>https://www.lena-nadir.dev/blog/ja/ai-103-exam-experience/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/ja/ai-103-exam-experience/</guid><description>Microsoft認定資格 AI-103（Azure AI Apps and Agents Developer Associate）にβ版で一発合格した体験記。出題範囲・配点比率、LLMを使ったハンズオン学習法、試験本番のTipsをまとめます</description><pubDate>Thu, 16 Jul 2026 00:00:00 GMT</pubDate></item><item><title>[AI-103] Generative AI/Agent領域 Hands-Onロードマップ</title><link>https://www.lena-nadir.dev/blog/ja/ai-103-series-00-roadmap/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/ja/ai-103-series-00-roadmap/</guid><description>AI-103のGenerative AI/Agent領域に絞った実践Hands-onシリーズの導入記事。シリーズ全体のテーマと構成をまとめます。</description><pubDate>Sat, 18 Jul 2026 00:00:00 GMT</pubDate></item><item><title>[AI-103 Hands-On] 第1回: Foundry PortalでAgentを作る</title><link>https://www.lena-nadir.dev/blog/ja/ai-103-series-01-agent-basics/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/ja/ai-103-series-01-agent-basics/</guid><description>Foundry PortalでWeb Search Toolを使うAgentをノーコードで作成し、Python SDK(Responses API)から同じAgentを呼び出すまでを、実機で一気通貫に確認します。</description><pubDate>Tue, 21 Jul 2026 00:00:00 GMT</pubDate></item><item><title>[AI-103 Hands-On] 第2回: File Searchで基本的なRAGを実装する</title><link>https://www.lena-nadir.dev/blog/ja/ai-103-series-02-file-search-rag/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/ja/ai-103-series-02-file-search-rag/</guid><description>Microsoft FoundryのFile Search機能でドキュメントをアップロードし、Agentがそのファイルをもとに検索拡張生成(RAG)で回答するまでを、Python SDK経由で実機で確認します。</description><pubDate>Sat, 25 Jul 2026 00:00:00 GMT</pubDate></item><item><title>[AI-103 Hands-On] 第3回: JSON構造化出力の方法</title><link>https://www.lena-nadir.dev/blog/ja/ai-103-series-03-structured-output/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/ja/ai-103-series-03-structured-output/</guid><description>Web Search Toolを持つ2つのFoundry Agent(Instructionsのみ、Instructions+JSON Schema)を作り、検索から構造化までを1回の呼び出しで完結させる方式を、パース成功率・型の安定性・トークンコストの面で実測比較します。</description><pubDate>Sat, 01 Aug 2026 00:00:00 GMT</pubDate></item><item><title>AI-500とは？試験範囲やAI-103との違い、他資格との比較</title><link>https://www.lena-nadir.dev/blog/ja/ai-500-beta-exam-trends/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/ja/ai-500-beta-exam-trends/</guid><description>Microsoftの新資格AI-500を公式情報から解説。試験範囲、AI-103との関係、難易度に関する初期報告、LangGraphとMicrosoft Agent Framework、AWS・Google Cloud・NVIDIA資格との違いを整理します。</description><pubDate>Sun, 16 Aug 2026 00:00:00 GMT</pubDate></item><item><title>主要AIエージェントフレームワーク &amp; プラットフォーム比較: LangChain, Microsoft, AWS, Googleの方向性や強み</title><link>https://www.lena-nadir.dev/blog/ja/ai-agent-framework-comparison/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/ja/ai-agent-framework-comparison/</guid><description>LangChain/LangGraph、Microsoft Agent Framework、AWS(Bedrock Agents/Strands Agents)、Google Agent Development Kit(ADK)の4つを、提供主体・得意領域・状態管理方式・相性の良いLLMの観点で整理して比較します。</description><pubDate>Sun, 09 Aug 2026 00:00:00 GMT</pubDate></item><item><title>データ基盤のモダナイゼーションとソリューション選定</title><link>https://www.lena-nadir.dev/blog/ja/data-platform-modernization/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/ja/data-platform-modernization/</guid><description>Microsoft Fabricで実務向けPoCを進めた後、Azure Databricksを再評価することになった事例から、データ基盤ソリューション選定で評価すべき要件、アーキテクチャの移植性、PoCの評価軸を整理します。</description><pubDate>Sat, 05 Sep 2026 00:00:00 GMT</pubDate></item><item><title>LangGraphでGenerator &amp; Critic Patternを実装 ― Agentが単体テスト仕様書を生成</title><link>https://www.lena-nadir.dev/blog/ja/langgraph-test-spec-generator/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/ja/langgraph-test-spec-generator/</guid><description>LangGraphのGenerator &amp; Critic Patternを使い、人間が作成した主要テスト項目を起点に、設計書とソースコードから単体テスト仕様書を補完・レビューする処理を実装する</description><pubDate>Sun, 30 Aug 2026 00:00:00 GMT</pubDate></item><item><title>Microsoft FoundryでParallel Fan-Out/Gather Patternを実装する ― 株式分析を題材にMulti-Agentを試す</title><link>https://www.lena-nadir.dev/blog/ja/multi-agent-stock-analysis/</link><guid isPermaLink="true">https://www.lena-nadir.dev/blog/ja/multi-agent-stock-analysis/</guid><description>Microsoft Foundryを使ってParallel Fan-Out/Gather Patternによるマルチエージェント構成を実装します。Google Developers Blogで紹介されているMulti-Agent Patternを整理し、株式分析をユースケースとして専門Agentの並列実行と結果統合を検証します。</description><pubDate>Wed, 26 Aug 2026 00:00:00 GMT</pubDate></item></channel></rss>