To match your experience to a job description with AI, paste the posting into a chatbot, have it list the requirements and sort them into must-haves and nice-to-haves, write your own evidence for each one, then ask the AI to rate every match as strong, partial or a gap, strictly. The result is an evidence map: which story proves which requirement, and where you'll need an honest answer about a gap. AI is good at close reading and strict grading. It can't supply evidence; a match that only exists in the AI's wording isn't a match.
Below: a sample evidence map, a seven-step workflow, five prompts, a worked answer to "Do you have experience with X?" and the checks that keep you from overclaiming. For a quick list of likely questions from any posting, the free interview questions from a job description tool does it in your browser. For the full method, see how to use AI for interview prep.
What an evidence map looks like
Here is a map we wrote for an operations coordinator posting, filled in by a retail assistant manager. The most useful column is the rating, because it tells you where you need a story and where you need a straight answer.
| Requirement | Must or nice | Your evidence | Match | Story to tell |
|---|---|---|---|---|
| Coordinate schedules across sites | Must | Built the weekly rota for two stores, about 40 staff | Strong | Holiday rota rebuild |
| Inventory systems | Must | Daily stock counts in the store system; no admin work | Partial | Reducing stock loss |
| Vendor management | Must | Deliveries and disputes with three suppliers | Partial | Late-delivery escalation |
| Advanced Excel | Nice | Pivot tables and lookups; no macros | Partial | Weekly sales tracker |
| SQL | Nice | None | Gap | Honest answer plus what you're doing about it |
What AI is good at here, and where it overclaims
| Use AI to | Don't let AI |
|---|---|
| List every requirement, including ones buried in paragraphs | Add requirements the posting doesn't contain |
| Separate must-haves from nice-to-haves by wording | Decide for you which gaps don't matter |
| Rate your evidence strictly | Rephrase your evidence so it sounds like a closer match |
| Predict questions for each requirement | Write answers with experience you don't have |
| Spot a requirement with no story | Fill that gap with a story from someone else's career |
Workflow: posting to evidence map
- Extract the requirements (prompt 1): every requirement as a short line, sorted into must-have and nice-to-have, plus the three the employer seems to stress most.
- Write your evidence yourself, one line per requirement, from memory and your résumé. If you're stuck, prompt 2 interviews you.
- Rate each pair strictly (prompt 3). When in doubt, the lower rating.
- Choose lead stories: one strong story for each must-have, starting with the three the employer stresses.
- Prepare partial-match answers (prompt 5): what's closest, how it differs, and what you're doing to close the difference.
- Predict the questions for your top requirements (prompt 4).
- Rehearse against the same posting in a mock interview built from it.
Five prompts to copy
1. Requirement extractor
Here is a job description: [paste]. List every requirement as a short line. Mark each as must-have or nice-to-have, based on the wording ("required", "preferred", "bonus") and where it appears. Then name the three requirements this employer seems to care about most and quote the words that suggest it. Don't add requirements that aren't in the text.
Why it works: quoting the words keeps the AI's guess about priorities checkable.
2. Evidence interview
For each must-have requirement below, ask me one question about when I've done it, one at a time. "Never" is an acceptable answer. Afterwards, list my evidence in my own words next to each requirement. Requirements: [paste]
Why it works: it gathers evidence without letting the AI phrase it for you.
3. Strict rater
Rate each requirement and evidence pair as strong (done directly, recently, with a result), partial (related or smaller-scale experience) or gap (nothing relevant). Be strict: when in doubt, choose the lower rating, and explain each rating in one line. Don't rephrase my evidence to make it sound closer. Pairs: [paste]
Why it works: an honest map is more useful than a flattering one, because the interviewer will grade it strictly too.
4. Likely questions per requirement
For the three most important requirements, write two questions an interviewer is likely to ask to test each one: one "tell me about a time" question and one direct "how would you" question.
Why it works: it turns the map into a practice list.
5. Partial-match answer
Help me outline an honest answer for this partial match: what I've done that's closest, how it's similar, how it's different, and what I'm doing to close the difference. Use only my facts. Requirement: [paste]. My evidence: [paste]
Why it works: "some, at a smaller scale" said first is far stronger than "yes" followed by a thin story.
Worked example: "Do you have vendor management experience?"
Weak first draft (example we wrote): "Yes, I have a lot of experience managing vendors. In my current role I work with vendors every day and I'm very comfortable with vendor relationships, negotiating and making sure everything runs smoothly."
What the strict rater said: partial. The evidence covers day-to-day supplier contact and disputes, not contracts or negotiation. Prompt 5 turned the evidence into an outline.
Improved answer (example we wrote): "Some, at a smaller scale than this role. As assistant store manager I was the day-to-day contact for our three main suppliers: I checked deliveries against orders, logged shortfalls and handled disputes. Last spring one supplier was late four weeks running, so I tracked every delivery for a month, took the record to their area rep and agreed a new delivery window that held. I haven't negotiated contracts; that sat with our regional office. The tracking habit is what I'd bring on day one, and the contract side is what I'd want to learn first."
The improved answer starts with "some". Claiming a full match on a partial one is the fastest way to stumble on the follow-up.
Checking the map before you rely on it
- Stretched matches. "Ran daily stand-ups" doesn't become "Agile project management experience" because the AI says so.
- Keyword stuffing. Use the employer's terms where they describe what you really did, not as decoration. Our guide to job description keywords in interview answers explains why stuffing backfires.
- Invented requirements. Compare the AI's list with the posting line by line.
- Misread priorities. The AI's "top three" is an inference. Confirm with the recruiter if you get the chance: "What matters most in the first few months?"
- Guessed tool names. If the posting names an internal or unfamiliar system, don't let the AI guess what it is.
- Privacy. The posting is public; your evidence may not be. Leave internal figures and client names out of the chat.
Rehearse against the same posting
Once the map is done, say each lead story out loud with the requirement in mind, and answer each partial-match question in under a minute. Then let something else pick the questions.
Our walkthrough of a mock interview from your résumé and job description shows how to set one up, and an AI mock interview built from the same posting checks whether the right story comes to mind under pressure. If a requirement turns into a résumé question, our guide on turning your résumé into interview answers covers it. In the interview itself, follow your employer's rules.
Where CoPilot Interview fits
CoPilot Interview is an AI interview assistant whose mock interviews start from your résumé and the job description, read questions aloud and score each spoken answer out of 10, which makes it a direct test of your evidence map. The free plan gives 15 minutes a day of live help in sessions of up to 15 minutes, with AI mock interviews included. Paid plans are Standard at $24.99 a month and Pro at $39.99 a month. It runs on a desktop app for Windows 10 and 11 and macOS 12 or later, a free iPhone and iPad companion on the App Store (iOS 16 or later) and an Android companion in beta.
For a no-install warm-up, the free practice demo has 145 questions across 10 tracks, scores typed answers and suggests a follow-up. In real interviews, follow your employer's rules.
Disclosure: CoPilot Interview is our product, so weigh what we say about it accordingly.
Frequently asked questions
How do I match my experience to a job description?
List the posting's requirements, write one line of real evidence for each, rate each match as strong, partial or a gap, and choose one story per must-have. AI speeds up the listing and the rating; the evidence has to be yours.
What if I only match some of the requirements?
That's normal; postings often list more than any one candidate has. Lead with your strong matches on the must-haves, answer partial matches honestly ("some, at a smaller scale"), and say what you're doing about any gap.
Is it OK to use AI to tailor my answers to a job description?
Yes, as long as the experience in the answers is real. Let AI map and grade; don't let it rephrase your evidence into a closer match than it is.
How many stories do I need to cover a job description?
Usually five or six: one per must-have requirement, with the three the employer stresses most covered first. One story can often prove two requirements.
Can AI tell me what the employer cares about most?
It can make a good guess from wording, order and repetition, and you should ask it to quote the words it used. Treat the guess as a hypothesis and confirm it with the recruiter if you can.
Rehearse it out loud before the real one
The AI mock interview builds questions from your resume and the job description, reads them aloud and scores each spoken answer out of 10. Included on the free plan.
Start a mock interview