Applying Is a Job in Itself

My first resume was a generic one. No keywords, no projects highlighted, nothing that would pass an ATS. Result: 60 applications, barely any responses.

From the few callbacks I got, one technical test. I wasn't prepared for it. The format, the content, the timing — everything caught me off guard. I failed it.

Before redoing the resume, I spent 3 to 4 months building personal projects. Learning the tech stack for the roles I was targeting — Docker, Terraform, Ansible, CI/CD — and practicing on real use cases. Those projects gave me something concrete to showcase, and more importantly, the right keywords to get past automated filters.

The change was immediate. On the next round of applications, 50% response rate. And of those responses, half were genuinely interested. The resume was finally doing its job.

Finding the Interview

Even with a better resume, landing an interview takes time. Weeks of outreach, follow-ups on LinkedIn, messages to people I didn't know. No network in the industry, no referrals, no connections.

Then Criteo responded. One LinkedIn application, among over 180 applicants. No referral, I didn't know anyone there. Just a resume that stood out.

10 Days to Learn

The interview was technical: algorithms, in Python. Two problems — I had never done serious algo work, and I didn't know Python. Yes, both at the same time.

I took 10 days to learn both in parallel. Python syntax, data structures, classic patterns — sorting, searching, recursion. Intense, but that's the kind of sprint 42 teaches you to handle.

Are 10 days enough? No. But I wasn't going to let the opportunity pass without trying.

The Interview

Day of. Video call, screen sharing on my end. And on the other side — no camera, no face. Just code on the screen. I don't even know if someone is there.

Result: I don't explain my reasoning. How do you walk through your logic when you don't know if anyone is listening? No reaction, no signal, no feedback. I write code in silence.

The questions: full Python programs, algorithmic problems to solve live. Nothing related to DevOps, Cloud, or infrastructure. The role was SRE. The test was pure algo.

For someone whose strength is explaining how they think, showing curiosity, putting choices in context — it's the worst possible format. I'm not saying it's a bad process. It fits what Criteo is looking for. But there's a gap between what this type of interview measures and what I have to offer. And that gap isn't about preparation — it's simply a format that doesn't suit me.

Failing an interview isn't failing, period. It's feedback on what to work on — and what to target better.

What I Took From It

Get better at coding. Algorithms aren't my strength, but it's a weakness I can reduce. Since that interview, I've been trying to practice more regularly, working on logic, pushing myself to solve problems. Not to become a LeetCode expert — but to stop being stuck when it comes up.

Target better. Not all companies hire the same way. Some swear by algorithms only. Others evaluate reasoning, projects, ability to learn, background. Criteo made me realize I needed to find companies whose process gives me a chance to show what I can actually do — not just what I can code in 45 minutes.

What Came Next

Failing an interview is feedback. It pushed me to work on things I would have kept avoiding, and to think about what I was really looking for.

What came next was an interview where the format matched — but that's another story.