Use AI to prepare a job application

Paste a résumé. Add a job posting. Ask AI to tailor it. You may get cleaner writing, but the model still knows only what you gave it.

Your résumé is a compressed record. It may leave out a difficult decision, a useful question you asked, or the reason a project changed direction. Those details can give you something specific to say.

This workflow produces five things: a role brief, an experience bank, an application argument, revised materials, and interview notes. Use it for a role worth investigating. It is a preparation method, not a guarantee of an interview.

1. Decide what needs checking first

Read the original posting. Record the role, level, location, responsibilities, stated requirements, and any application instructions. Separate explicit requirements from your interpretation. A job title alone tells you little about the actual work.

For international students, keep work-authorization and sponsorship questions visible. Use our employer-research guide and recruiter-question builder when the wording is unclear. A persuasive application does not change an employer’s stated policy.

Make a practical decision: continue, ask a question, build a missing skill, or focus elsewhere. AI can organize the information, but it cannot know an unpublished hiring rule or decide your authorization for you.

2. Build a short role brief

Use the employer’s site, product information, and current posting. Note three things:

• What work does the posting explicitly describe?
• What might make that work difficult? Label this as a hypothesis.
• What would you need to ask a recruiter or someone familiar with the role?

Keep a source and date for company facts. If AI supplies a link or quotation, open it and check that it supports the claim. Do not turn a plausible guess into “I know your team is struggling with…” in a cover letter.

Try this prompt:

Using only the source excerpts below, summarize the role’s stated responsibilities. Separate verified details from your hypotheses. Identify two useful questions I could ask to understand the work better. Do not claim to have visited a link or know the team’s internal priorities. Ask for missing information.

3. Interview yourself before rewriting yourself

Choose one rough experience. It can be a course project, research, volunteering, part-time work, or work abroad. Explain it in ordinary language.

Ask AI to explore it before producing résumé language:

Ask me one question at a time about this experience. Explore what was difficult, what I was responsible for, what I personally decided and did, and what happened. Distinguish my work from the team’s. Do not propose achievements or numbers for me to accept. After at most eight questions, organize the facts as confirmed, uncertain, and missing. Ask for my approval before drafting anything.

An experience does not need a dramatic outcome. A documented handoff, a corrected error, or an honest lesson can be useful if it addresses the work. Do not add a percentage simply because a résumé example includes one.

4. Put evidence beside the requirement

Open the application evidence planner. Choose up to three requirements and record a supporting example plus any unresolved question.

Fictional example: a role asks for checking data quality. A student remembers preparing a survey file. Further questioning reveals that they noticed repeat entries, confirmed the duplicate-handling rule with a supervisor, applied it, and recorded the changes.

“Assisted with data analysis” now has supporting detail. A possible factual bullet is: “Checked duplicate survey entries against a supervisor-approved rule and documented changes to the research file.” No performance metric is implied.

If another requirement asks for production SQL experience and the student has none, record that gap. A class exercise might demonstrate some SQL knowledge; it should not be renamed production experience.

5. Choose the argument before the wording

Write one sentence explaining why your experience is relevant to this role. Support it with one or two examples. This is your working argument, not necessarily a sentence to paste into the application.

For the fictional student: “I have practice checking research data carefully and making my changes reviewable.” That is narrower and easier to substantiate than “I am an exceptional analytical leader.”

Use that argument to choose what deserves space in the résumé. For a cover letter, develop an example that adds context rather than reciting every bullet. See Write a cover letter with something to say.

Editing prompt:

Propose changes using only my approved facts and the role’s stated requirements. Show the original, proposed edit, supporting fact, and reason. Preserve titles, dates, seniority, and ownership. Ask when evidence is missing. Do not add skills or estimate results. Flag sentences that could fit almost any applicant.

Harvard’s résumé guidance emphasizes specific, factual language. Its AI guidance treats AI as an aid to improving authentic material. Our workflow applies those principles through an explicit fact-review step.

6. Make international experience understandable

A U.S. reader may not recognize an organization, title, or qualification. Add brief factual context when needed. For example, explain what an organization does rather than calling it “the equivalent of” a famous American company without evidence.

Keep original credentials accurate. Do not invent a U.S. degree equivalence or convert a grade without an authoritative method. If the application requests a credential evaluation, follow its instructions.

Ask AI what a reader might need explained, then supply the actual context. The model should help you clarify what happened, not change it to resemble an American experience.

7. Rehearse the questions your application creates

Read each claim and ask: “Could I explain this without the document in front of me?” Practice what you did, why you chose it, what happened, and what remains uncertain.

Use STAR-L for past-experience questions: Situation, Task, Action, Result, and Link to the role. Ask AI follow-up questions after your own first attempt. A script is less useful if you cannot explain its reasoning.

If you need a skill you have not demonstrated, use AI for learning and practice. A small, clearly labeled project using public or fictional data may help you explore it. It does not replace a credential or experience requirement, and you should not claim a practice project was paid work.

Before you hit apply

• Follow the employer’s instructions on format, attachments, and AI assistance.
• Confirm names, dates, links, job title, and company references.
• Check that every skill and outcome has evidence behind it.
• Remove confidential information from anything you share with an AI service.
• Read the final wording aloud. Keep language you understand and can explain.
• Save the submitted version and the original posting for interview preparation.

The free workspace organizes your notes locally and builds a prompt you can review and take to an AI tool. It does not analyze your résumé, score your fit, or send an application. Use a career adviser or trusted reviewer for another perspective.