How to Tailor Your Resume to a Job Description with AI (15 min)

Here’s the single biggest mistake I see job seekers make: they spend an hour picking which roles to apply to, then send the exact same CV to every one.

I can spot this in 10 seconds. When I’m reviewing 40 applications for a product manager role and 37 of them have the same generic summary (“results-driven professional with cross-functional experience”), the 3 that reference our actual product get prioritized. That’s not bias — that’s rational triage. The three tailored ones look like they want this job. The 37 look like they want any job.

AI makes tailoring fast enough that you have no excuse not to do it. This article is the exact workflow I give candidates I’m coaching. 15 minutes per application. With practice, 10. It’s fast because you’re not rewriting everything — you’re surgically editing the parts that matter.

The 60-second test: does your CV look tailored?

Before the workflow, here’s the diagnostic. Open your current CV. Look at:

1. Your summary/headline (first 2-3 lines): Does it reference something specific about this job title? Or is it generic (“Product manager with 6 years experience…”)?
2. Your top 3 bullets: Do they include keywords from the job description? Or are they the same bullets you wrote 2 years ago?
3. Your skills section: Does it show the skills this specific role asks for in the order the JD lists them?

If all three are generic, you’re mass-applying. Tailoring means these three things change per application.

The 7-step workflow (15 minutes)

Step 1: Parse the job description (2 min)

Don’t read the JD normally. Extract structure. Paste this into ChatGPT or Claude:

Extract from this job description:
- Top 5 HARD skills (tools, technologies, certifications)
- Top 3 SOFT skills (explicitly mentioned, not inferred)
- Any must-have qualifications (degree, years of experience, location)
- The core problem this role solves for the company

Output as 4 short lists. Do not add anything not in the JD.

JD:
[paste full JD]

Output example for a product manager job description:

Hard skills: SQL, Amplitude, Jira, Figma, roadmap documentation Soft skills: stakeholder management, written communication, decision-making under ambiguity Must-haves: 5+ years PM experience, B2B SaaS background Core problem: reducing time-to-value for new customers

Keep this visible while you work. It’s your scorecard.

Step 2: Map your experience to each requirement (3 min)

For each requirement from step 1, find the single best bullet in your existing CV that demonstrates it. Some will have strong matches, some weak, some none.

Example mapping:

You now know: the two strong matches need to go front and center. The medium match needs rewriting. The gap (Amplitude) — you either add a related tool you know, or skip it.

Step 3: Pick 3-5 bullets to rewrite (1 min)

This is the mistake most people make: they try to rewrite the whole CV for each application. Don’t. You’ll burn out and the rewrites will get worse as you tire.

Pick 3-5 bullets maximum:

• 2-3 that map to the top requirements from step 1
• 1-2 that add keywords the CV currently lacks

Mark them. Everything else stays.

Step 4: Use AI to draft new versions (5 min)

This is where you use the prompts from my ChatGPT prompts for resume guide. The one I use most is the Bullet Rewriter:

You are a senior recruiter who reads 300 CVs a week. Rewrite this bullet point
to be specific, metric-led, and memorable. Emphasize: [paste relevant requirement
from step 1]. Keep under 22 words. Do not use: leveraged, spearheaded, results-driven,
dynamic, passionate, cross-functional, synergistic, robust, holistic, deliverables.

Original bullet:
[paste]

Role I'm applying to: [paste 2-sentence summary of the JD]

Output: 3 variations, briefly explain which keyword from the JD each emphasizes.

Example output for the Metabase/SQL case:

1. “Built 10+ SQL-driven dashboards in Metabase, reducing ad-hoc data requests by 60%.” — emphasizes SQL from JD
2. “Shipped 10+ self-serve dashboards that cut data-request tickets from 8/week to 3/week.” — emphasizes data accessibility
3. “Wrote 40+ SQL queries to power product analytics dashboards used weekly by 12 stakeholders.” — emphasizes SQL depth + stakeholder scope

Pick the one that matches the role’s emphasis. In this case, variation 3 — the JD wanted SQL + stakeholders.

Step 5: Edit in your voice (3 min)

AI drafts need editing. Always. Run the bullets you picked through this checklist:

• Strip any of the 13 buzzwords from this list
• Verify every number — did you actually write 40 queries? If you wrote 20, say 20
• Read aloud — does it sound like you describing your job to a friend? Or like a LinkedIn motivational post?
• Cut anything that claims rather than shows — “proven”, “passionate”, “dedicated” are all claims

Usually you’ll end up with bullets that are 70% AI-drafted and 30% your edits. That’s the right ratio.

Step 6: Update the summary / headline (1 min)

This is the highest-leverage 1 minute in the whole workflow. Your CV’s top 2-3 lines decide whether I keep reading.

Default format I’ve seen work:

[Job title from their JD] with [X years] experience in [specific thing their JD cares about]. Recently [most relevant concrete achievement, 1 metric]. [What you’d like to do next, matching their role].

Example for the PM role above:

Product Manager with 6 years in B2B SaaS. Recently cut customer onboarding time 40% via a self-serve revamp. Looking to apply that playbook to [company]‘s time-to-value problem.

3 lines. Mentions their title, their industry, their actual problem. That beats a generic 5-line summary every time. (If you’re working from a UK base format, the UK CV format 2026 guide covers what stays fixed and what gets retailored per application.)

Step 7: Run the final check (1 min)

Before sending, look at the first 8 seconds of your CV — the time a recruiter will spend on it.

That’s roughly: your name, your summary, and your first 2 bullets.

Do those first 4-5 items collectively mention:

• The role title (or close variant)?
• Their industry (if specific)?
• At least 2 of the top skills from step 1?

If yes, ship. If no, the summary and first bullet need another pass.

Real before-and-after

Here’s an anonymized example from a candidate I helped place — a product designer applying to a fintech startup.

Before (generic CV, sent to 20 roles):

Results-driven product designer with a proven track record of leveraging cross-functional synergy to deliver impactful user experiences across dynamic, fast-paced environments.

• Leveraged user research to inform design decisions across multiple products
• Spearheaded cross-functional initiatives to improve product outcomes
• Built strong relationships with engineering and product teams

After (tailored to the fintech role):

Product Designer, 5 years in consumer finance. Shipped 3 zero-to-one features at [prev company] including a savings product used by 40k+ users. Looking to design financial products that feel safe without being boring.

• Designed and shipped savings account onboarding that improved signup-to-funded conversion from 38% to 61% in 6 weeks
• Partnered weekly with engineering (4 ICs) and compliance (2 leads) to ship KYC flows meeting FinCEN requirements
• Led user research for card issuance product — 22 customer interviews, synthesized into the v2 design spec

The before version is atmosphere. The after version is a person. Same role, same candidate, completely different signal to the recruiter reading it.

The after version got her 4 interviews in the next 10 days. The before version had gotten 2 in 6 weeks.

When tools like Teal and Rezi help (and when they don’t)

The workflow above uses ChatGPT or Claude. That’s fine — they’re flexible, they work for free, you learn what good tailoring looks like.

Dedicated tools like Teal are useful when:

• You’re applying to 10+ roles a month — the Chrome extension auto-saves JDs, tracker keeps you organized
• You want the JD-to-CV keyword match quantified automatically
• You’re tired of pasting the same prompts into ChatGPT

Teal is essentially the workflow above, built into an app. You can use either approach. The workflow is the thing that matters.

Rezi is similar but optimized more heavily for ATS keyword density. Better if your main problem is getting filtered out by specific companies’ ATS systems. I break down when to use which in Teal vs Rezi.

Both tools still require the human edit step (step 5). Neither produces ship-ready output unedited — I’ve seen both tools’ raw output, and recruiters can spot it.

Common mistakes candidates make

Rewriting every bullet. You tire out, the later ones are sloppy, and you actually need the unchanged “baseline” bullets to show continuity.

Not updating the summary. People will tailor 3 bullets but leave their generic summary in place. The summary is the first thing read — tailor it first.

Pasting the whole JD into one AI prompt. Too much context leads to generic output. Extract the specific requirements first (step 1), then use them surgically.

Inventing metrics. If your bullet said “improved engagement” and AI suggests “improved engagement by 43%”, check whether 43% is real. If it isn’t, use the real number or remove the metric entirely. Fake metrics get caught in interviews.

Tailoring for roles you don’t actually want. If you’re not excited about the role, the tailoring will be mediocre. Save your tailoring energy for roles you’d say yes to.

The math on this

15 minutes per application. 4 applications a day = 1 hour. That’s 20 tailored applications per week.

Compare to the untailored alternative: 50 spray-and-pray applications per week, 0.5% callback rate = 1 interview every 2-3 weeks.

My placed candidates who did tailored applications averaged 1 interview per 10 applications. So 20 tailored = 2 interviews per week. That’s 10x better than untailored, at a fraction of the time.

The workflow works. Do it 15 times this week. You’ll feel the difference by Friday.

Quick references

• Prompts to use: ChatGPT prompts for resume
• Buzzwords to strip from AI output: 13 AI resume buzzwords recruiters hate
• Tools that automate this workflow: Teal vs Rezi comparison
• Main resume pillar for everything AI + CV

Tailoring isn’t optional anymore. AI removed the excuse. The candidates who do it consistently — 15 minutes per role, 3-5 applications per day — land faster than the ones spraying 50/week. Be the first kind.

Related reading

• How to start a cover letter: 5 openers that work — the letter that sits next to the tailored CV.
• Cover letter for a job you’re underqualified for — the 3-paragraph structure when tailoring isn’t enough.
• LinkedIn skills to add in 2026 — the skill-filter layer that works parallel to CV keyword tailoring.
• Transferable skills for career change — for candidates whose current CV doesn’t map cleanly to the target job.

Tools that automate parts of this workflow

• Rezi review — builds ATS-keyword-matched CVs from pasted job descriptions.
• Jobscan review — scores your tailored CV against the target job before you submit.
• Teal review — the broader workflow if you’re tailoring 10+ per week.