<?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: nicky pappo</title>
    <description>The latest articles on DEV Community by nicky pappo (@nicky_pappo_bae454101c10e).</description>
    <link>https://dev.to/nicky_pappo_bae454101c10e</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%2F3938194%2F7c802f10-46da-4700-8915-35cfb4255dfa.png</url>
      <title>DEV Community: nicky pappo</title>
      <link>https://dev.to/nicky_pappo_bae454101c10e</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/nicky_pappo_bae454101c10e"/>
    <language>en</language>
    <item>
      <title>Hello DEV community – building Screen DLP (AI that detects phone cameras aimed at screens)</title>
      <dc:creator>nicky pappo</dc:creator>
      <pubDate>Mon, 18 May 2026 13:19:10 +0000</pubDate>
      <link>https://dev.to/nicky_pappo_bae454101c10e/hello-dev-community-building-screen-dlp-ai-that-detects-phone-cameras-aimed-at-screens-393k</link>
      <guid>https://dev.to/nicky_pappo_bae454101c10e/hello-dev-community-building-screen-dlp-ai-that-detects-phone-cameras-aimed-at-screens-393k</guid>
      <description>&lt;p&gt;Hey everyone, I'm Nicky, CEO of ScreenStop.&lt;/p&gt;

&lt;p&gt;We're building Screen DLP — a new security category that traditional&lt;br&gt;
  DLP completely misses.&lt;/p&gt;

&lt;p&gt;records, trading positions, or source code and photographs it with&lt;br&gt;
  their phone. The data never moves digitally — so network monitoring,&lt;br&gt;
  clipboard controls, and file restrictions are all blind to it.&lt;/p&gt;

&lt;p&gt;What we built: on-device AI running on the endpoint's webcam that&lt;br&gt;
  detects when a phone is in capture position — and blurs the screen&lt;br&gt;
  before the photo is taken. No cloud, no data leaves the machine.&lt;/p&gt;

&lt;p&gt;The hard engineering problem was false positive rate. Hands, coffee&lt;br&gt;
  cups, second monitors — all trigger naive models. Most of our work&lt;br&gt;
  went into making it not annoying to use.&lt;/p&gt;

&lt;p&gt;Pre-seed, currently in pilots in healthcare and financial services.&lt;/p&gt;

&lt;p&gt;Happy to discuss the detection model, the architecture, or the&lt;br&gt;
  category itself. What's your take — is this a gap you've seen in&lt;br&gt;
  your own security stack?&lt;/p&gt;

&lt;p&gt;&lt;a href="https://screenstop.co" rel="noopener noreferrer"&gt;https://screenstop.co&lt;/a&gt;&lt;/p&gt;

&lt;p&gt;Tags: security, ai, machinelearning, privacy, DLP&lt;/p&gt;

</description>
      <category>ai</category>
      <category>security</category>
      <category>showdev</category>
      <category>startup</category>
    </item>
  </channel>
</rss>
