As featured in
Using AI to sharpen your job search - and where humans still decide.
By the numbers
% of firms already auto-screen applications — Jobcloud
% run a chatbot first interview — Jobcloud
applicants on a single posting, often
In the Beobachter, our co-founder Klaus Fuchs explains how AI now sits at the front of most hiring - and how candidates can turn that to their advantage. His read is pragmatic: AI makes applying faster for everyone, and the people who understand how it reads them get ahead.
The single biggest lever is staying close to the job ad: use the same keywords and wording as the listing, and adapt your CV for each posting rather than blasting the same one everywhere. Old tricks like hidden white-text keywords are dead - the screening models flag unusual keyword density.
Because AI scans endless material, you can rarely submit too much proof - certificates, diplomas and references all count. Motivation letters matter less since ChatGPT; they're only worth it with a genuinely personal connection to the company or role.
And it stays human. The later rounds are still people meeting people - a shared connection, a former internship, someone inside the company. So invest in real networking. Our own takeaway: work with a recruiter who gets you past the automated first filter to a person, and knows the roles that never reach a job board.
“AI makes applying faster for everyone - the people who understand how it reads them get ahead.”
What this means
The read for both sides of the market.
For companies
- Screen fairly, decide humanly. Automation handles volume, but the deciding rounds still need real people.
- Make the brief legible. Clear, keyword-honest postings get you better-matched applicants from the machines doing the first pass.
- The best never apply. Many strong candidates move through trusted networks, not job boards - meet them there.
For candidates
- Tailor every CV. Mirror the ad's keywords and wording; a generic CV loses at the automated first gate.
- Pile on the proof. Certificates, diplomas and references all count - the machine can read all of it.
- Network for the human rounds. A connection, an event, a recruiter who knows you - that's where decisions are still made.