<?xml version="1.0" encoding="utf-8" standalone="yes"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
  <channel>
    <title>Edge Computing on App Coding</title>
    <link>https://appcoding.com/tags/edge-computing/</link>
    <description>Recent content in Edge Computing on App Coding</description>
    <generator>Hugo</generator>
    <language>en-us</language>
    <lastBuildDate>Mon, 05 Oct 2026 00:00:00 +0000</lastBuildDate>
    <atom:link href="https://appcoding.com/tags/edge-computing/index.xml" rel="self" type="application/rss+xml" />
    <item>
      <title>A Feature Store Small Enough to Embed: Point-in-Time Reads for Local and Edge ML</title>
      <link>https://appcoding.com/a-feature-store-small-enough-to-embed-point-in-time-reads-for-local-and-edge-ml/</link>
      <pubDate>Mon, 05 Oct 2026 00:00:00 +0000</pubDate>
      <guid>https://appcoding.com/a-feature-store-small-enough-to-embed-point-in-time-reads-for-local-and-edge-ml/</guid>
      <description>&lt;p&gt;A churn model looks excellent in the notebook and mediocre in production. The model isn&amp;rsquo;t broken. The training table computed &lt;code&gt;orders_7d&lt;/code&gt; with a rolling window over the full history, so some rows counted orders placed after the label date. The serving code, written later in another language, counts a calendar week instead of seven days. One feature, two definitions, and a quiet look into the future. Those are the two bugs a feature store exists to prevent: training-serving skew and label leakage.&lt;/p&gt;</description>
    </item>
  </channel>
</rss>
