<?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: Mohamed Wazeem</title>
    <description>The latest articles on DEV Community by Mohamed Wazeem (@wazcodes).</description>
    <link>https://dev.to/wazcodes</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.us-east-2.amazonaws.com%2Fuploads%2Fuser%2Fprofile_image%2F2740110%2Fa96f9bb2-054c-49a5-8c6a-15997f6e2bd8.png</url>
      <title>DEV Community: Mohamed Wazeem</title>
      <link>https://dev.to/wazcodes</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/wazcodes"/>
    <language>en</language>
    <item>
      <title>What is Jev?</title>
      <dc:creator>Mohamed Wazeem</dc:creator>
      <pubDate>Fri, 25 Sep 2026 19:20:26 +0000</pubDate>
      <link>https://dev.to/wazcodes/what-is-jev-1ope</link>
      <guid>https://dev.to/wazcodes/what-is-jev-1ope</guid>
      <description>&lt;p&gt;Jev is the hottest AI topic this week.&lt;/p&gt;

&lt;p&gt;What is Jev btw?&lt;/p&gt;

&lt;p&gt;Simply put, it is a decision AI model used for decision-making.&lt;/p&gt;

&lt;p&gt;Unlike LLMs (which predict the next token in a sequence and generate text based on that), Jev by TypeSafe AI, the first System One Model, is designed to make decisions based on these question types:&lt;/p&gt;

&lt;ol&gt;
&lt;li&gt;Choice — selects an option and returns it with probabilities and confidence.&lt;/li&gt;
&lt;li&gt;Score — rates content and returns a score, a probability for each level, and confidence.&lt;/li&gt;
&lt;li&gt;Noul — evaluates Yes/No questions and returns the probability for yes.&lt;/li&gt;
&lt;/ol&gt;

&lt;p&gt;Jev isn’t designed to replace LLMs; it complements them. We can combine LLMs and Jev to execute functions based on decisions made by Jev.&lt;/p&gt;

&lt;p&gt;The most interesting part is that, according to TypeSafe’s published evaluations, Jev is 193.6× faster and 444.6× cheaper than the compared LLM setup.&lt;/p&gt;

&lt;p&gt;Jev is currently priced at $42 per 1 billion input tokens.&lt;/p&gt;

&lt;p&gt;That’s why it’s getting attention.&lt;/p&gt;

&lt;p&gt;Currently, new signups are paused, but Jev is accessible through the Vercel AI SDK.&lt;/p&gt;

&lt;p&gt;For learning more about Jev, check out: &lt;a href="//docs.typesafe.ai"&gt;docs.typesafe.ai&lt;/a&gt;&lt;/p&gt;

</description>
      <category>ai</category>
      <category>jev</category>
      <category>llm</category>
    </item>
    <item>
      <title>What I Learned While Installing Python</title>
      <dc:creator>Mohamed Wazeem</dc:creator>
      <pubDate>Sat, 01 Nov 2025 10:00:07 +0000</pubDate>
      <link>https://dev.to/wazcodes/what-i-learned-while-installing-python-2o5o</link>
      <guid>https://dev.to/wazcodes/what-i-learned-while-installing-python-2o5o</guid>
      <description>&lt;h2&gt;
  
  
  How to Install Python
&lt;/h2&gt;

&lt;ol&gt;
&lt;li&gt;Go to &lt;a href="https://www.python.org/downloads/" rel="noopener noreferrer"&gt;python.org/downloads&lt;/a&gt;.&lt;/li&gt;
&lt;li&gt;Download the Python installer based on your operating system.&lt;/li&gt;
&lt;li&gt;Install Python using the downloaded file.

&lt;ul&gt;
&lt;li&gt;Make sure you check “Add Python to PATH” if you are using Windows.&lt;/li&gt;
&lt;/ul&gt;
&lt;/li&gt;
&lt;/ol&gt;

&lt;h2&gt;
  
  
  What I Found Interesting
&lt;/h2&gt;

&lt;p&gt;Since my school days, I’ve installed Python many times and even helped others do it. But only recently did I realize an important factor during installation — adding Python to the environment variables.&lt;/p&gt;

&lt;p&gt;I had noticed that typing &lt;code&gt;python&lt;/code&gt; in the Command Prompt didn’t work; instead, I had to use &lt;code&gt;py&lt;/code&gt; to check the version of Python I was using. I used to think that &lt;code&gt;python&lt;/code&gt; and &lt;code&gt;pip&lt;/code&gt; worked only for versions before 3.10, and that for later versions we had to use &lt;code&gt;py&lt;/code&gt; and &lt;code&gt;pip3&lt;/code&gt;.&lt;/p&gt;

&lt;p&gt;But I was wrong. The real reason was that I hadn’t added Python to the PATH environment variable during installation.&lt;/p&gt;

</description>
      <category>python</category>
      <category>programming</category>
      <category>learning</category>
    </item>
  </channel>
</rss>
