Boost Your Job Search Productivity with This AI-Driven System

Flowchart depicting an AI-driven job search workflow for product designers. The process starts with Bardeen scraping job listings, which are imported into a Google Sheets file. Google Sheets uses custom filters via App Scripts, and OpenAI (GPT-40 mini) analyzes job descriptions to filter relevant offers. The final document contains job listings that meet the user’s criteria, streamlining the job search process.

Job hunting takes time and effort. From building a design portfolio to crafting a CV and completing interviews and tasks, the process can feel overwhelming.

I am looking for a product design job, and I have been thinking about ways to streamline the process using AI and automation.

To save time, I developed an AI system to filter jobs quickly and efficiently, boosting my productivity.

In this article, I’ll show you how I built it so you can create your own and save time in your job search.

First, let’s look at how the system works

The system quickly scrapes job listings and pastes them into a Google Sheets file.

In less than a minute, I run a custom filters, built with App Script using ChatGPT, which filters out jobs that don’t match my interests. Then, using OpenAI’s API, the remaining job offers are summarized.

The Google Sheet includes four sections:

  • New Jobs: Recently scrapped job listings.
  • In Process: Jobs I’ve applied to and am waiting on a response.
  • Not Interested: Jobs I’ve decided not to apply for.
  • Rejected: Jobs that reviewed my application and declined.
The image shows a job application tracker spreadsheet in Google Sheets. The spreadsheet contains columns labeled “Company Name,” “Job Title,” “Job Description,” “Date Posted,” “Application Link,” “Application Status,” “Location,” “Start analyze,” and “Short Summary.” There is an arrow pointing to the tab titled “New jobs,” indicating that this tab is selected. The spreadsheet lists job details, mostly for remote roles across different locations such as Belgium, Malta, and Estonia. The “Applicati
Job application tracker spreadsheet

Here’s the process:

  1. Scrapes job listings and copies them into a Google Sheets file.
  2. Built-in App Scripts handle the first round of filtering.
  3. ChatGPT reads and analyzes job descriptions.
  4. I manually review the offers.
  5. I apply only to jobs that pass all filters.
The image is a flowchart outlining an AI-driven job search process. Here’s a breakdown of the steps: 1. Scrape job listings and paste them into Google Sheets. 2. Filter job listings. 3. AI summarizes job offers. 4. Manual review of job offers. 5. Decision point: “Am I interested in this job?” If yes, move to review step: “Do I want to apply to this job?” 6. If yes, proceed to “Apply to jobs”; if not, move the job to the “Not Interested” category.
AI-driven job search process flowchart

Scrapping job listing website and copy to a Google Sheets file with Bardeen

Start by collecting job listings from multiple sites. The goal is to collect as many offers as possible in a Google Sheets file.

Visit job sites like Wellfound and Glassdoor. Search for roles you’re interested in, such as remote product designer positions.

Then, use Bardeen (a tool that automates browser tasks and scrapes web data) to scrape key details like company name, job description, and more.

The free version of Bardeen is enough for this task. Watch this video for a step-by-step guide. The process is simple and quick to learn.

Pro tip: Save job listing pages with filters applied to easily check for new offers later.

The GIF shows an automation tool being used to scrape job listings from a website.
Scrapping process

Some scripts in the Google Sheets file handle the first round of filtering

After collecting job listings, I use Google Script in Google Sheets to create custom filters.

Google Script is a feature that extends and automates Google Workspace apps. No coding skills are needed. Just ask ChatGPT to generate the required functions.

Watch this video for a quick tutorial on adding Google Scripts to Google Sheets.

For example, I asked it to create scripts that:

Remove Applied Jobs
If a job I applied for appears in the sheet, the script deletes it.

Remove Duplicate Jobs
If a job appears more than once in the sheet, the script removes the duplicate (common when scraping from multiple pages).

Remove Duplicate Jobs in the File
Jobs I’m not interested in, applied to, or rejected are deleted automatically. The script checks for duplicates between new and existing jobs, deleting any new listings automatically.

The GIF shows an action in a Google Sheets document, specifically filtering job listings in a spreadsheet titled “Job_Applications_Tracker.” The filtering appears to be happening in real-time as the user interacts with the sheet. The same columns from earlier, including “Company Name,” “Job Title,” “Date Posted,” “Application Link,” and “Location,” are visible. The user selects and navigates the rows of job postings, with a focus on remote job listings in various countries. Tabs at the bottom of
Filtering job listings

GPT-40 mini analyzes the job description and automates the filtering process

After setting up the initial filters (which were straightforward thanks to App Scripts), I needed an advanced filter.

My goal was for AI to read job descriptions and filter them based on specific parameters.

I automated this process using Pabbly, a tool similar to Zapier.

Here’s how it works: Every time I check a box in a Google Sheets cell, the automation triggers. It sends the job description to GPT-40 mini, which analyzes the details based on my criteria. The AI pulls out three key pieces of information:

  1. What the company does (so I can avoid certain industries).
  2. Whether the position is remote, hybrid, or on-site.
  3. If the job listing includes salary details.

This allows me to quickly decide if the job is a fit. For example, if the salary is too low or the position is on-site when I prefer remote, I can move the job to my “Not Interested” sheet without reading the entire description.

The setup was simple. First, I connected Google Sheets to Pabbly. Then, I wrote a prompt that instructs the LLM model on how to filter the job descriptions. The hardest part was refining the prompt to meet my exact needs. After a few iterations, I got it right, and now the process runs smoothly.

The GIF shows a Google Sheets document being used to analyze job descriptions in real-time. The spreadsheet, titled “Job_Applications_Tracker,” features rows with various job postings. A few rows in the “Short Summary” column show text beginning with “### Summarized,” and checkboxes in the “Start analyze” column appear to indicate which rows are being processed or analyzed. The job locations range from “Remote” to places like Catalonia, Spain. Tabs at the bottom still show categories like “New j
GPT-40 mini analyzes the job description

I apply only to job offers that pass all filters

After filtering, I open each job and review the details. If I lose interest, it goes to the “Not Interested” sheet. If the job fits, I apply and move it to the “In Process” sheet.

To conclude

In this blog post, I showed you how I built an AI-driven workflow to streamline the job search process for product designers.

The workflow combines tools like Bardeen, Google Sheets, App Scripts, and an LLM model (GPT-40 mini) to automate tasks such as scraping job listings, filtering relevant offers, and summarizing descriptions.

Here’s how it works:

Job scraping: Bardeen automatically collects job listings from sites like Wellfound and Glassdoor.

Filtering: Google Sheets and App Scripts filter out duplicates and irrelevant jobs. The LLM model reviews job descriptions based on criteria like industry, work format, and salary.

Decision-making: After filtering, I apply only to jobs that meet the requirements, saving time and effort.

This process simplifies the job search, letting you focus on applying to the right roles faster.

👉 BTW, I’m looking for my next product design role in complex or AI-driven products. If you’re hiring, feel free to reach out via email and connect with me on LinkedIn.

Thank you for reading the article. Please feel free to share it with your friends or team members, and if you have any questions, please let me know.

If you enjoyed my article, I suggest you follow me so you’ll receive an email whenever I post.

You can also follow me on Linkedin, where I share tips several times a week.

Automate 90% of Your Design Job Search with This AI Workflow was originally published in UX Planet on Medium, where people are continuing the conversation by highlighting and responding to this story.