<?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: Invoxy</title>
    <description>The latest articles on DEV Community by Invoxy (@invoxy).</description>
    <link>https://dev.to/invoxy</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%2F3724982%2F399f2c1a-debf-4ffd-9c14-72b043198dea.png</url>
      <title>DEV Community: Invoxy</title>
      <link>https://dev.to/invoxy</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/invoxy"/>
    <language>en</language>
    <item>
      <title>How to Practice for a Technical Interview with your own Resume (Free AI Tool)</title>
      <dc:creator>Invoxy</dc:creator>
      <pubDate>Thu, 22 Jan 2026 02:09:45 +0000</pubDate>
      <link>https://dev.to/invoxy/i-built-a-resume-interview-bot-because-chatgpt-was-too-generic-12f</link>
      <guid>https://dev.to/invoxy/i-built-a-resume-interview-bot-because-chatgpt-was-too-generic-12f</guid>
      <description>&lt;p&gt;I've been prepping for engineering interviews lately, and I noticed a problem.&lt;/p&gt;

&lt;p&gt;If I ask ChatGPT: &lt;em&gt;"Roast my resume,"&lt;/em&gt; it gives generic advice like "Add more metrics."&lt;br&gt;
If I ask: &lt;em&gt;"Simulate a technical interview,"&lt;/em&gt; it asks standard questions like "What is the difference between TCP and UDP?"&lt;/p&gt;

&lt;p&gt;The missing link was &lt;strong&gt;Context&lt;/strong&gt;. The LLM didn't know &lt;em&gt;my&lt;/em&gt; specific history.&lt;/p&gt;

&lt;p&gt;So, I built a tool to fix it.&lt;/p&gt;

&lt;h3&gt;
  
  
  The Project: Invoxy
&lt;/h3&gt;

&lt;p&gt;I built &lt;strong&gt;&lt;a href="https://invoxy.xyz" rel="noopener noreferrer"&gt;Invoxy.xyz&lt;/a&gt;&lt;/strong&gt;. It’s a free tool that parses your PDF resume and generates context-aware interview questions.&lt;/p&gt;

&lt;h3&gt;
  
  
  How it works under the hood
&lt;/h3&gt;

&lt;ol&gt;
&lt;li&gt;
&lt;strong&gt;Client-Side Parsing:&lt;/strong&gt; I used &lt;code&gt;pdf-parse&lt;/code&gt; to extract raw text strings from the resume upload directly in the browser (keeping data private).&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Context Injection:&lt;/strong&gt; I built a system prompt that feeds the LLM your specific bullet points + the target Job Description.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;The "Red Team" Persona:&lt;/strong&gt; I tuned the model to be skeptical. Instead of being nice, it specifically looks for gaps in employment or vague claims (e.g., &lt;em&gt;"You claimed 20% growth here—how exactly did you attribute that to your code?"&lt;/em&gt;).&lt;/li&gt;
&lt;/ol&gt;

&lt;h3&gt;
  
  
  The Stack
&lt;/h3&gt;

&lt;ul&gt;
&lt;li&gt;  &lt;strong&gt;Frontend:&lt;/strong&gt; Next.js&lt;/li&gt;
&lt;li&gt;  &lt;strong&gt;AI:&lt;/strong&gt; Google API (Context-aware prompting)&lt;/li&gt;
&lt;li&gt;  &lt;strong&gt;Parsing:&lt;/strong&gt; PDF.js / text-extraction&lt;/li&gt;
&lt;/ul&gt;

&lt;h3&gt;
  
  
  Try it out (Free Beta)
&lt;/h3&gt;

&lt;p&gt;I released it today as a free beta. If you are job hunting, give it a shot and let me know if the "Red Team" mode is too mean.&lt;/p&gt;

&lt;p&gt;&lt;strong&gt;Live Demo:&lt;/strong&gt; &lt;a href="https://invoxy.xyz" rel="noopener noreferrer"&gt;https://invoxy.xyz&lt;/a&gt;&lt;/p&gt;

</description>
      <category>webdev</category>
      <category>ai</category>
      <category>javascript</category>
    </item>
  </channel>
</rss>
