I wanted to test something. Could AI completely automate the job search process?
So I built a system:
- Claude generates tailored cover letters
- A script auto-fills applications
- AI customizes my resume for each role
- Everything tracked in a Notion database
I applied to 200 developer jobs in one week.
The Numbers
| Metric | Count |
|---|---|
| Applications sent | 200 |
| Auto-rejected (within 24h) | 127 |
| Ghosted (no response in 2 weeks) | 48 |
| Screening calls | 18 |
| Technical interviews | 7 |
| Final rounds | 3 |
| Offers | 1 |
0.5% conversion rate. One offer from 200 applications.
What I Learned
1. ATS systems are more sophisticated than you think
63% of my applications were auto-rejected. Why? Because companies now use AI to detect AI-generated applications. Fighting AI with AI is an arms race you can't win.
2. The cover letters didn't matter
Of the 18 screening calls I got, only 2 recruiters mentioned my cover letter. Most admitted they never read them. The resume and portfolio did all the heavy lifting.
3. Quantity vs Quality is a false choice
The one offer I received? It came from an application where I spent 45 minutes researching the company and writing a genuine, specific message to the hiring manager on LinkedIn.
Not from the 199 AI-blasted applications.
4. The hidden job market is real
During this experiment, I also got 2 interview invitations from:
- A comment I left on a Dev.to article
- A DM conversation on Twitter about a technical topic
Neither required a formal application.
The Strategy That Actually Works
Stop mass-applying. Start targeted outreach.
Week 1-2: Identify 10 companies you genuinely want to work for
Week 3-4: Research each company deeply:
- Read their engineering blog
- Use their product
- Find pain points you could solve
Week 5-6: Make contact:
- Comment on their team's posts
- Contribute to their open source
- Send a specific, value-adding message to a team lead
Week 7-8: Apply formally with a portfolio that shows you understand their problems
This approach has a 10-15% conversion rate vs 0.5% for mass applications.
The Controversial Take
AI is making the job market worse for people who use it lazily, and better for people who use it strategically.
Don't use AI to apply faster. Use it to prepare better:
- Research companies faster
- Practice interview questions
- Build portfolio projects that solve real problems
- Write better documentation for your projects
The job market isn't broken. Our approach to it is.
More career strategies and AI tips for developers: SwiftUI Daily on Telegram
Top comments (0)