<?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>DEV Community: Satyam Sharma</title>
    <description>The latest articles on DEV Community by Satyam Sharma (@satyamskillz).</description>
    <link>https://dev.to/satyamskillz</link>
    <image>
      <url>https://media2.dev.to/dynamic/image/width=90,height=90,fit=cover,gravity=auto,format=auto/https:%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F717535%2Fa75f1045-49c3-4aea-9cf8-9fef6da2fb0c.jpg</url>
      <title>DEV Community: Satyam Sharma</title>
      <link>https://dev.to/satyamskillz</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/satyamskillz"/>
    <language>en</language>
    <item>
      <title>Building a System That Automates YouTube Post-Production</title>
      <dc:creator>Satyam Sharma</dc:creator>
      <pubDate>Mon, 25 May 2026 18:53:55 +0000</pubDate>
      <link>https://dev.to/satyamskillz/building-a-system-that-automates-youtube-post-production-55d</link>
      <guid>https://dev.to/satyamskillz/building-a-system-that-automates-youtube-post-production-55d</guid>
      <description>&lt;p&gt;Most creator tools today focus on analytics and suggestions.&lt;/p&gt;

&lt;p&gt;We wanted to explore something different:&lt;/p&gt;

&lt;p&gt;What if YouTube post-production could actually be automated?&lt;/p&gt;

&lt;p&gt;We’ve been building Growati, a system that helps creators generate personalised:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;titles&lt;/li&gt;
&lt;li&gt;descriptions&lt;/li&gt;
&lt;li&gt;thumbnails&lt;/li&gt;
&lt;li&gt;chapters&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;for YouTube videos.&lt;/p&gt;

&lt;p&gt;One of the most difficult engineering problems was properly handling video understanding.&lt;/p&gt;

&lt;p&gt;Instead of treating videos like plain text, our system analyses videos frame-by-frame to extract:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;key moments&lt;/li&gt;
&lt;li&gt;best frames&lt;/li&gt;
&lt;li&gt;visual context&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The goal is to make metadata generation more contextual instead of generic AI text generation.&lt;/p&gt;

&lt;p&gt;Another challenge was orchestration.&lt;/p&gt;

&lt;p&gt;A single automation flow may involve:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;frame extraction&lt;/li&gt;
&lt;li&gt;metadata generation&lt;/li&gt;
&lt;li&gt;thumbnail generation&lt;/li&gt;
&lt;li&gt;scoring&lt;/li&gt;
&lt;li&gt;updates&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;We ended up using a queue-based architecture with BullMQ to handle automation reliably.&lt;/p&gt;

&lt;p&gt;&lt;a href="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fhqxjqv6twuqciadlmj6l.png" class="article-body-image-wrapper"&gt;&lt;img src="https://media2.dev.to/dynamic/image/width=800%2Cheight=%2Cfit=scale-down%2Cgravity=auto%2Cformat=auto/https%3A%2F%2Fdev-to-uploads.s3.amazonaws.com%2Fuploads%2Farticles%2Fhqxjqv6twuqciadlmj6l.png" alt=" " width="800" height="450"&gt;&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;One thing we learned quickly while talking to creators:&lt;/p&gt;

&lt;p&gt;Many educational creators and podcasters are overwhelmed by post-production work after uploading.&lt;/p&gt;

&lt;p&gt;Most of them are not looking for more dashboards.&lt;/p&gt;

&lt;p&gt;They want less manual work.&lt;/p&gt;

&lt;p&gt;That insight changed how we approached the product.&lt;/p&gt;

&lt;p&gt;Right now, Growati is still early:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;10+ beta users&lt;/li&gt;
&lt;li&gt;500+ synced videos&lt;/li&gt;
&lt;li&gt;50+ metadata updates processed&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;We’re launching on Product Hunt on 28 May.&lt;/p&gt;

</description>
      <category>webdev</category>
      <category>ai</category>
      <category>startup</category>
      <category>youtube</category>
    </item>
  </channel>
</rss>
