Getting Hired in the AI Age

Part 1

Getting through the filter

Applicant tracking software reads your résumé into a database. Anything it cannot read is invisible, and 'invisible' looks identical to 'not qualified'.

1. 1

   One column, always

   Two-column layouts are the most common reason a good résumé parses as nonsense — the software reads straight across and interleaves the columns. Sidebars for skills and dates are the worst offender. This alone sinks a large number of applications.

2. 2

   No tables, text boxes, headers or footers

   Content inside these is frequently dropped entirely. People put their phone number in the header and get told the application was incomplete.

3. 3

   Use the headings the software expects

   Experience, Education, Skills. Not 'Where I've Made an Impact'. Creative headings are a real reason for content to land in the wrong field or nowhere.

4. 4

   Use their words

   If the posting says 'accounts receivable', do not write 'AR' and assume it matches. If it says 'customer success', do not rely on 'customer service' being counted. Mirror the exact phrasing where it is honestly true of you.

5. 5

   Spell out and abbreviate both

   'Certified Public Accountant (CPA)' matches either search. Cheap, and it costs one line.

6. 6

   Send .docx unless told otherwise

   PDFs mostly parse now, but .docx remains the safest. Never send a scan or an image — it is unreadable to the software and you will never know.

7. 7

   Put dates in a normal format

   'March 2021 – June 2024'. Unusual formats produce gaps in the parsed record that look like unexplained unemployment.

Part 2

Using AI on your application, properly

Employers assume AI was involved. What they are actually screening for is whether you can back up what it wrote.

1. 1

   Tailor, do not generate

   Paste your real résumé and the job description and ask which of your genuine experience is most relevant and what vocabulary to mirror. That is editing with a research assistant. Generating a history is a different thing and it ends in an interview you cannot survive.

2. 2

   Never claim a skill you do not have

   AI will happily add anything to a résumé. A technical interview finds it in ten minutes, and 'I put it on there but have not used it' is worse than never claiming it.

3. 3

   Keep your own voice

   Fully AI-written cover letters read the same as each other and recruiters see hundreds. Write badly in your own words first, then ask for tightening. That order matters and produces a noticeably better result.

4. 4

   Ask it to interrogate you

   'Ask me questions about this job until you have enough to help me write about my experience.' Far better output than asking for a letter cold, because it surfaces things you would not have thought to mention.

5. 5

   Get an honest read on the posting

   Paste it and ask what this employer actually seems to want, what is likely negotiable, and what the phrasing suggests about the team. Surprisingly good at reading between lines.

6. 6

   Be able to defend every line

   The test before sending: could you talk for two minutes about each bullet point? If not, cut it.

Part 3

Recorded and AI-assisted interviews

One-way video interviews and automated screening calls are now common, especially for high-volume roles. They feel strange and there are specific things that help.

• You usually get the question, short preparation time, and one or two takes. Read the instructions before starting — the rules vary and cannot be undone.
• Look at the camera lens, not at yourself on screen. It reads as eye contact and it is the single biggest improvement available.
• Light from in front, not behind. A window facing you beats any equipment.
• Answer in a clear structure: the situation, what you did, what happened. Recorded formats punish rambling more than live ones because there is no interviewer to rescue you.
• Say the keywords out loud. Many systems transcribe and search the text of what you said.
• Test the camera and microphone on the actual platform beforehand, not on a different app.
• Practice once with a recording of yourself. It is uncomfortable and it is the fastest improvement you will make.
• If you need an accommodation, ask in writing before the assessment, not after.

Part 4

What is now worth more, not less

AI has changed what is scarce. Worth aiming at deliberately rather than assuming the old advice still holds.

• Being able to verify. Anyone can produce a draft; the person who can tell whether it is right is the one worth employing.
• Judgment about what should be done, not just how to do it.
• Anything physical. Trades, care, repair, installation — these did not become easier to automate and demand is rising.
• Trust and relationships. Customers still want to deal with a person they know.
• Clear writing and speaking. More output exists, so being understood quickly is worth more.
• Knowing the tools. Being the person on a team who actually uses them well is a real differentiator right now.
• Domain knowledge, because that is what tells you when the confident answer is wrong.

Part 5

Job scams, which have got much better

AI made fraudulent listings fluent and fake recruiters convincing. Job seekers are targeted because urgency and hope make people less careful.

• No legitimate employer asks for money — for training, equipment, a background check, or a starter kit.
• No legitimate employer needs your bank details or Social Security number before a written offer.
• Interviews conducted entirely over text chat, especially on Telegram or WhatsApp, are a strong signal.
• A check to deposit and partially forward is always fraud, and the bank will reclaim the whole amount from you.
• Verify the company independently: find their real website yourself, call the main number, confirm the person exists.
• An offer with no interview, well above market pay, is not luck.
• Look up the recruiter on LinkedIn separately. Brand-new profiles with few connections are worth pausing over.