<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>TemJayto</title>
    <link>https://www.temjayto.com/</link>
    <description>Field notes on Fabric, Power BI, Copilot and Azure AI agents, written from real customer work.</description>
    <language>en</language>
    <lastBuildDate>Wed, 07 Oct 2026 00:00:00 GMT</lastBuildDate>
    <atom:link href="https://www.temjayto.com/feed.xml" rel="self" type="application/rss+xml"/>
    <image>
      <url>https://www.temjayto.com/assets/og-image.png</url>
      <title>TemJayto</title>
      <link>https://www.temjayto.com/</link>
    </image>
    <item>
      <title>From Weeks to Hours: Engineering an AI Pipeline for Site Investigation Reports (complete article)</title>
      <link>https://www.temjayto.com/site-report-ai-pipeline.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/site-report-ai-pipeline.html</guid>
      <pubDate>Wed, 07 Oct 2026 00:00:00 GMT</pubDate>
      <category>Azure AI &amp; Agents</category>
      <description>The whole story on one page: why a liability document can't be a chatbot, hosted agents on Azure AI Foundry, every fact on a cited deterministic path, and an engineer holding the pen.</description>
    </item>
    <item>
      <title>Part 1: From Weeks to Hours, Why a Site Investigation Report Is Not a Chatbot Problem</title>
      <link>https://www.temjayto.com/site-report-part1-business-problem.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/site-report-part1-business-problem.html</guid>
      <pubDate>Wed, 07 Oct 2026 00:00:00 GMT</pubDate>
      <category>Azure AI &amp; Agents</category>
      <description>What a Site Investigation Report is, why it takes an engineer weeks, why "point an LLM at it" is wrong, and the end-to-end pipeline that replaces the copy-paste.</description>
    </item>
    <item>
      <title>Part 2: The Model Narrates, Code Decides</title>
      <link>https://www.temjayto.com/site-report-part2-model-narrates-code-decides.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/site-report-part2-model-narrates-code-decides.html</guid>
      <pubDate>Wed, 07 Oct 2026 00:00:00 GMT</pubDate>
      <category>Azure AI &amp; Agents</category>
      <description>The boundary the whole design defends: provenance on every value, a reason on every gap, no collapsed parcels, conflicts as divergences, and a machine-checked summary.</description>
    </item>
    <item>
      <title>Part 3: Hosted Agents on Azure AI Foundry</title>
      <link>https://www.temjayto.com/site-report-part3-hosted-agents-foundry.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/site-report-part3-hosted-agents-foundry.html</guid>
      <pubDate>Wed, 07 Oct 2026 00:00:00 GMT</pubDate>
      <category>Azure AI &amp; Agents</category>
      <description>Per-agent Entra identities, an attach-only Site object every hop must pass, two deliberate human pauses, and provenance that doubles as the OpenTelemetry trace.</description>
    </item>
    <item>
      <title>Part 4: The Data Layer, Authoritative Sources Wrapped as Guarded Tools</title>
      <link>https://www.temjayto.com/site-report-part4-guarded-data-layer.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/site-report-part4-guarded-data-layer.html</guid>
      <pubDate>Wed, 07 Oct 2026 00:00:00 GMT</pubDate>
      <category>Azure AI &amp; Agents</category>
      <description>Regrid, FEMA, USDA, wetlands and Planetary Computer behind guarded tools, plus an ordinance corpus in Azure AI Search that can only cite the parcel's own jurisdiction.</description>
    </item>
    <item>
      <title>Part 5 Finale: Trust Engineering, Business Impact and Lessons for Agentic Document Pipelines</title>
      <link>https://www.temjayto.com/site-report-part5-trust-and-lessons.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/site-report-part5-trust-and-lessons.html</guid>
      <pubDate>Wed, 07 Oct 2026 00:00:00 GMT</pubDate>
      <category>Azure AI &amp; Agents</category>
      <description>The finale: offline tests, runtime gates that halt instead of degrade, production evaluation, the business economics, six lessons, and where the roadmap goes next.</description>
    </item>
    <item>
      <title>From Raw Data to Stunning Report: the complete Lakehouse to Power BI series on one page</title>
      <link>https://www.temjayto.com/lakehouse-to-beautiful-report.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/lakehouse-to-beautiful-report.html</guid>
      <pubDate>Wed, 16 Sep 2026 00:00:00 GMT</pubDate>
      <category>Fabric &amp; AI</category>
      <description>The whole journey on one page: Fabric MCP loads bronze, Copilot engineers gold, Power BI MCP builds the semantic model and report, and Figma MCP makes it beautiful.</description>
    </item>
    <item>
      <title>Part 5 Finale: Make It Beautiful with Figma + Copilot</title>
      <link>https://www.temjayto.com/article5-beautiful-figma-report.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/article5-beautiful-figma-report.html</guid>
      <pubDate>Wed, 16 Sep 2026 00:00:00 GMT</pubDate>
      <category>Design &amp; AI</category>
      <description>The payoff: Figma MCP pulls your design tokens, Copilot themes the report, layout matches Figma frames. Ship the beautiful outcome.</description>
    </item>
    <item>
      <title>Part 4: Build the Report Locally with Copilot</title>
      <link>https://www.temjayto.com/article4-build-report.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/article4-build-report.html</guid>
      <pubDate>Tue, 15 Sep 2026 00:00:00 GMT</pubDate>
      <category>Power BI &amp; AI</category>
      <description>PBIP in VS Code on your Part 3 model: Copilot generates pages, binds visuals, wires interactions. Functional first, beauty next.</description>
    </item>
    <item>
      <title>Part 3: Semantic Models - Why They Matter + Build with Copilot</title>
      <link>https://www.temjayto.com/article3-semantic-models.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/article3-semantic-models.html</guid>
      <pubDate>Mon, 14 Sep 2026 00:00:00 GMT</pubDate>
      <category>Power BI &amp; AI</category>
      <description>Why models beat raw tables: one definition of sales, DirectLake speed, DAX measures + RLS, all generated via Power BI MCP.</description>
    </item>
    <item>
      <title>Part 2: Data Engineering Bronze to Gold with Copilot</title>
      <link>https://www.temjayto.com/article2-data-engineering.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/article2-data-engineering.html</guid>
      <pubDate>Sun, 13 Sep 2026 00:00:00 GMT</pubDate>
      <category>Fabric &amp; AI</category>
      <description>Copilot-written Spark notebooks clean bronze into silver, quarantine bad rows, and build gold star-schema tables ready for modeling.</description>
    </item>
    <item>
      <title>Part 1: Load Data into the Lakehouse with Copilot + Fabric MCP</title>
      <link>https://www.temjayto.com/article1-lakehouse-loading.html</link>
      <guid isPermaLink="true">https://www.temjayto.com/article1-lakehouse-loading.html</guid>
      <pubDate>Sat, 12 Sep 2026 00:00:00 GMT</pubDate>
      <category>Fabric &amp; AI</category>
      <description>Start where every good report starts: raw files to OneLake bronze. Set up Fabric MCP, create shortcuts, and let Copilot build your ingestion pipeline.</description>
    </item>
  </channel>
</rss>
