<?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: Rowharbor Data</title>
    <description>The latest articles on DEV Community by Rowharbor Data (@rowharbor).</description>
    <link>https://dev.to/rowharbor</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%2F4242905%2F75bef10e-f318-42cb-8ad3-b62be6eb92bd.png</url>
      <title>DEV Community: Rowharbor Data</title>
      <link>https://dev.to/rowharbor</link>
    </image>
    <atom:link rel="self" type="application/rss+xml" href="https://dev.to/feed/rowharbor"/>
    <language>en</language>
    <item>
      <title>What I learned pulling 1,650 Gulf jobs from five different ATS feeds</title>
      <dc:creator>Rowharbor Data</dc:creator>
      <pubDate>Fri, 13 Nov 2026 12:37:58 +0000</pubDate>
      <link>https://dev.to/rowharbor/what-i-learned-pulling-1650-gulf-jobs-from-five-different-ats-feeds-29ba</link>
      <guid>https://dev.to/rowharbor/what-i-learned-pulling-1650-gulf-jobs-from-five-different-ats-feeds-29ba</guid>
      <description>&lt;p&gt;I wanted a simple thing: every open job in the UAE, Saudi Arabia and the rest of the Gulf, straight from the companies hiring, without job-board reposts.&lt;/p&gt;

&lt;p&gt;It turns out most companies don't post jobs by hand. Their careers page runs on an applicant tracking system (ATS), and the big ones publish &lt;strong&gt;public JSON feeds&lt;/strong&gt; of every open role. No login, no scraping HTML, no proxies. The feeds exist so careers pages can render, and anyone can read them.&lt;/p&gt;

&lt;p&gt;I ended up reading feeds from 48 employers across five systems and got about 1,650 Gulf jobs per run. Here's what I learned.&lt;/p&gt;

&lt;h2&gt;
  
  
  1. Every ATS has a feed, and every feed is different
&lt;/h2&gt;

&lt;div class="table-wrapper-paragraph"&gt;&lt;table&gt;
&lt;thead&gt;
&lt;tr&gt;
&lt;th&gt;ATS&lt;/th&gt;
&lt;th&gt;Endpoint (public)&lt;/th&gt;
&lt;th&gt;Quirk&lt;/th&gt;
&lt;/tr&gt;
&lt;/thead&gt;
&lt;tbody&gt;
&lt;tr&gt;
&lt;td&gt;Greenhouse&lt;/td&gt;
&lt;td&gt;&lt;code&gt;GET boards-api.greenhouse.io/v1/boards/{company}/jobs?content=true&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Descriptions are HTML, sometimes escaped &lt;strong&gt;twice&lt;/strong&gt;
&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Lever&lt;/td&gt;
&lt;td&gt;&lt;code&gt;GET api.lever.co/v0/postings/{company}?mode=json&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Dates are Unix milliseconds; extra locations hide in &lt;code&gt;categories.allLocations&lt;/code&gt;
&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Ashby&lt;/td&gt;
&lt;td&gt;&lt;code&gt;GET api.ashbyhq.com/posting-api/job-board/{company}?includeCompensation=true&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Salary is nested in &lt;code&gt;compensation.summaryComponents&lt;/code&gt;
&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;Workday&lt;/td&gt;
&lt;td&gt;&lt;code&gt;POST {host}/wday/cxs/{tenant}/{site}/jobs&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Search API, paginated by 20; dates are text like "Posted 3 Days Ago"&lt;/td&gt;
&lt;/tr&gt;
&lt;tr&gt;
&lt;td&gt;SmartRecruiters&lt;/td&gt;
&lt;td&gt;&lt;code&gt;GET api.smartrecruiters.com/v1/companies/{id}/postings?country=ae&lt;/code&gt;&lt;/td&gt;
&lt;td&gt;Has a country filter (nice!), but descriptions need a second call per job&lt;/td&gt;
&lt;/tr&gt;
&lt;/tbody&gt;
&lt;/table&gt;&lt;/div&gt;

&lt;p&gt;The cheapest one to work with is Greenhouse: one request, every job, full descriptions.&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight python"&gt;&lt;code&gt;&lt;span class="k"&gt;import&lt;/span&gt; &lt;span class="n"&gt;httpx&lt;/span&gt;

&lt;span class="k"&gt;def&lt;/span&gt; &lt;span class="nf"&gt;greenhouse_jobs&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;company&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt; &lt;span class="kt"&gt;str&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="o"&gt;-&amp;gt;&lt;/span&gt; &lt;span class="kt"&gt;list&lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="kt"&gt;dict&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="n"&gt;url&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;f&lt;/span&gt;&lt;span class="s2"&gt;"https://boards-api.greenhouse.io/v1/boards/{company}/jobs?content=true"&lt;/span&gt;
    &lt;span class="n"&gt;r&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;httpx&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;get&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;url&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;timeout&lt;/span&gt;&lt;span class="o"&gt;=&lt;/span&gt;&lt;span class="mi"&gt;30&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
    &lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;raise_for_status&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
    &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;json&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"jobs"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;

&lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;job&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="nf"&gt;greenhouse_jobs&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"careem"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="k"&gt;print&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;job&lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"title"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"|"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;job&lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"location"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="s2"&gt;"name"&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2&gt;
  
  
  2. "Is this job in the Gulf?" is harder than it looks
&lt;/h2&gt;

&lt;p&gt;Location fields are free text: &lt;code&gt;Dubai&lt;/code&gt;, &lt;code&gt;Dubai, UAE&lt;/code&gt;, &lt;code&gt;Remote - EMEA&lt;/code&gt;, &lt;code&gt;Riyadh, Saudi Arabia; London&lt;/code&gt;. My first matcher was a list of substrings: &lt;code&gt;dubai&lt;/code&gt;, &lt;code&gt;riyadh&lt;/code&gt;, &lt;code&gt;qatar&lt;/code&gt;, &lt;code&gt;oman&lt;/code&gt;…&lt;/p&gt;

&lt;p&gt;Then a job in &lt;strong&gt;Romania&lt;/strong&gt; showed up as Oman.&lt;/p&gt;

&lt;p&gt;"R-&lt;strong&gt;oman&lt;/strong&gt;-ia." Substring matching on short country names is a trap. The fix was word boundaries:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight python"&gt;&lt;code&gt;&lt;span class="k"&gt;import&lt;/span&gt; &lt;span class="n"&gt;re&lt;/span&gt;

&lt;span class="no"&gt;GULF&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;
    &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="s2"&gt;"\bdubai\b"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"Dubai"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"United Arab Emirates"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="s2"&gt;"\babu ?dhabi\b"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"Abu Dhabi"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"United Arab Emirates"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="s2"&gt;"\buae\b|united arab emirates"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="kc"&gt;None&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"United Arab Emirates"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="s2"&gt;"\briyadh\b"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"Riyadh"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"Saudi Arabia"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="s2"&gt;"\bdoha\b|\bqatar\b"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="kc"&gt;None&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"Qatar"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
    &lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="s2"&gt;"\bmuscat\b|\boman\b"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="kc"&gt;None&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;"Oman"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt;
&lt;span class="p"&gt;]&lt;/span&gt;
&lt;span class="no"&gt;GULF_RE&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="p"&gt;[&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;re&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nb"&gt;compile&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;p&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;re&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="no"&gt;I&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;city&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;country&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt; &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;p&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;city&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;country&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="no"&gt;GULF&lt;/span&gt;&lt;span class="p"&gt;]&lt;/span&gt;

&lt;span class="k"&gt;def&lt;/span&gt; &lt;span class="nf"&gt;gulf_match&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;text&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;rx&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;city&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;country&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="no"&gt;GULF_RE&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
        &lt;span class="k"&gt;if&lt;/span&gt; &lt;span class="n"&gt;rx&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;search&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;text&lt;/span&gt; &lt;span class="ow"&gt;or&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
            &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="n"&gt;city&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;country&lt;/span&gt;
    &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="kc"&gt;None&lt;/span&gt;

&lt;span class="nf"&gt;gulf_match&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="s2"&gt;"Bucharest, Romania"&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;  &lt;span class="c1"&gt;# None, finally&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;p&gt;Also check &lt;strong&gt;all&lt;/strong&gt; locations, not just the first. Plenty of roles list London first and Dubai third.&lt;/p&gt;

&lt;h2&gt;
  
  
  3. HTML that was escaped twice
&lt;/h2&gt;

&lt;p&gt;Some Greenhouse descriptions come back as &lt;code&gt;&amp;amp;lt;p&amp;amp;gt;About us&amp;amp;lt;/p&amp;amp;gt;&lt;/code&gt;, and a few were escaped &lt;em&gt;again&lt;/em&gt; (&lt;code&gt;&amp;amp;amp;lt;p&amp;amp;amp;gt;&lt;/code&gt;). One &lt;code&gt;html.unescape()&lt;/code&gt; isn't enough. I unescape in a loop until the text stops changing, then strip tags:&lt;br&gt;
&lt;/p&gt;

&lt;div class="highlight js-code-highlight"&gt;
&lt;pre class="highlight python"&gt;&lt;code&gt;&lt;span class="k"&gt;import&lt;/span&gt; &lt;span class="n"&gt;html&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;re&lt;/span&gt;

&lt;span class="k"&gt;def&lt;/span&gt; &lt;span class="nf"&gt;clean&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;raw&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
    &lt;span class="n"&gt;text&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;raw&lt;/span&gt; &lt;span class="ow"&gt;or&lt;/span&gt; &lt;span class="s2"&gt;""&lt;/span&gt;
    &lt;span class="k"&gt;for&lt;/span&gt; &lt;span class="n"&gt;_&lt;/span&gt; &lt;span class="k"&gt;in&lt;/span&gt; &lt;span class="nb"&gt;range&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="mi"&gt;3&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
        &lt;span class="n"&gt;un&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;html&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;unescape&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;text&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
        &lt;span class="k"&gt;if&lt;/span&gt; &lt;span class="n"&gt;un&lt;/span&gt; &lt;span class="o"&gt;==&lt;/span&gt; &lt;span class="n"&gt;text&lt;/span&gt;&lt;span class="p"&gt;:&lt;/span&gt;
            &lt;span class="k"&gt;break&lt;/span&gt;
        &lt;span class="n"&gt;text&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;un&lt;/span&gt;
    &lt;span class="n"&gt;text&lt;/span&gt; &lt;span class="o"&gt;=&lt;/span&gt; &lt;span class="n"&gt;re&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;sub&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="s2"&gt;"&amp;lt;[^&amp;gt;]+&amp;gt;"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;" "&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;text&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
    &lt;span class="k"&gt;return&lt;/span&gt; &lt;span class="n"&gt;re&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;sub&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="n"&gt;r&lt;/span&gt;&lt;span class="s2"&gt;"\s+"&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="s2"&gt;" "&lt;/span&gt;&lt;span class="p"&gt;,&lt;/span&gt; &lt;span class="n"&gt;text&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;&lt;span class="p"&gt;.&lt;/span&gt;&lt;span class="nf"&gt;strip&lt;/span&gt;&lt;span class="p"&gt;(&lt;/span&gt;&lt;span class="p"&gt;)&lt;/span&gt;
&lt;/code&gt;&lt;/pre&gt;

&lt;/div&gt;



&lt;h2&gt;
  
  
  4. Workday doesn't give you dates
&lt;/h2&gt;

&lt;p&gt;Workday returns &lt;code&gt;"postedOn": "Posted 30+ Days Ago"&lt;/code&gt;. You can turn it into an approximate date, but you can't get better than "about a month ago". Sort by it, but don't promise exact dates.&lt;/p&gt;

&lt;h2&gt;
  
  
  5. Extracting signals from job text: the false positives teach you the most
&lt;/h2&gt;

&lt;p&gt;Gulf job seekers care about things job boards rarely expose: visa sponsorship, whether Arabic is required, and whether a role is &lt;strong&gt;reserved for nationals&lt;/strong&gt; (Emiratisation and Saudization roles). I extracted these with plain regex rules rather than an AI model, then checked about 165 real postings by hand. The mistakes were the interesting part:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;"UAE Nationals Preferred" is not "UAE Nationals only."&lt;/strong&gt; My first rule flagged both. Preferred roles are still open to expats, so I added a negative lookahead for &lt;em&gt;preferred / desirable / a plus&lt;/em&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;Not every nationals tag is in English.&lt;/strong&gt; Some AECOM titles say "إماراتيين" ("Emiratis"), and Saudi graduate roles say "Tamheer" (a nationals training programme). Both needed their own patterns.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;"Arabic cuisine" is not a language requirement.&lt;/strong&gt; A Sous Chef ad was flagged "Arabic required". The fix: &lt;code&gt;\barabic\b(?!\s+(?:cuisine|food|coffee))&lt;/code&gt;.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;"Over 25 years of experience"&lt;/strong&gt; is usually the &lt;em&gt;company&lt;/em&gt; bragging, not the job requirement. I ignore numbers above 20, and when an ad mentions several, I keep the &lt;strong&gt;smallest&lt;/strong&gt;. If you're filtering for junior roles, you'd rather see a borderline job than miss one.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;"Spark" isn't always Apache Spark.&lt;/strong&gt; Hotel sales ads say "spark conversations". I now require context like "Apache Spark", "PySpark" or Spark appearing in a list of tools.&lt;/li&gt;
&lt;/ul&gt;

&lt;p&gt;The general lesson: rules are cheap and explainable, but you only find their holes by reading real data. When a rule isn't sure, return &lt;code&gt;null&lt;/code&gt; rather than guess.&lt;/p&gt;

&lt;h2&gt;
  
  
  6. Be polite, and only fetch what you keep
&lt;/h2&gt;

&lt;p&gt;Two things kept my runs fast and friendly to the servers:&lt;/p&gt;

&lt;ul&gt;
&lt;li&gt;
&lt;strong&gt;List first, fetch details later.&lt;/strong&gt; For Workday and SmartRecruiters, descriptions cost one extra request per job. I fetch the list, filter by location and keywords, sort newest first, and only then fetch details for jobs I actually keep, a few at a time with a small delay.&lt;/li&gt;
&lt;li&gt;
&lt;strong&gt;A real User-Agent, a short delay, and backoff on &lt;code&gt;429&lt;/code&gt;.&lt;/strong&gt; With those in place I haven't been rate-limited by any of these feeds.&lt;/li&gt;
&lt;/ul&gt;

&lt;h2&gt;
  
  
  7. The numbers
&lt;/h2&gt;

&lt;p&gt;For one full run across 48 employers: about &lt;strong&gt;1,650 Gulf jobs&lt;/strong&gt;, mostly hospitality (Accor alone had 821), engineering (AECOM 320) and delivery (Delivery Hero 129), plus a long tail of tech and fintech. Salaries were published for only a handful of jobs. Most Gulf employers still don't post pay.&lt;/p&gt;




&lt;p&gt;If you want to skip building this yourself, I packaged it as an &lt;a href="https://apify.com/rowharbor/gulf-employer-jobs" rel="noopener noreferrer"&gt;Apify Actor&lt;/a&gt; ($1.50 per 1,000 jobs). It's my own project, and the code examples above work on their own. Questions or ATS quirks you've hit? Drop them in the comments.&lt;/p&gt;

</description>
      <category>python</category>
      <category>webscraping</category>
      <category>datascience</category>
      <category>career</category>
    </item>
  </channel>
</rss>
