AI Resume Builder and Portfolio Maker
Explore templates and pricing
Which sections a resume needs and in what order, how to write bullets a recruiter can check, tailoring for each job ad, ATS basics, and the mistakes to cut.
Canonical: https://www.prezumi.com/blog/how-to-write-a-resume
A resume is a one-page document built from five core sections in a set order: a header with your name and contact details, a two-to-three-sentence summary, work experience in reverse-chronological order, education, and skills. Optional sections — certifications, languages, projects, interests — earn a place when they add evidence. Recruiters and screening software both expect that structure, which means most resume questions have a boring, reliable answer, and this guide collects them in one place.
It works as an overview: the rules here apply to nearly every resume, and where a situation needs more depth — a first job with no experience, a student CV, an academic CV with publications, a promotion at the same company, Latin honors, or using ChatGPT on the draft — there's a dedicated guide linked in context.
Two readers decide your resume's fate, and they read differently. The first is usually software: most employers run applications through an applicant tracking system before a person sees them, and Harvard Business School's Hidden Workers study found that more than 90% of the employers it surveyed use such systems to filter or rank candidates. The second is a human who, by the widely cited estimate, gives a first scan roughly seven seconds before deciding whether to keep reading.
Every rule in this guide follows from those two facts. The software needs clean, single-column text with standard section headings so it can extract your details into the right fields. The human needs the strongest, most specific evidence near the top, because that's all the first scan touches.
Name, phone number, a sensible email address, your city, and a LinkedIn URL if the profile is filled in. That's the whole section. Skip the photo, date of birth, and full street address in most markets — recruiters don't need them and often prefer not to receive them. If you're applying for jobs in the EU, some employers ask for the standardized Europass CV format instead, which has its own header conventions.
Two or three sentences: who you are, the role you're applying for, and the most concrete evidence you can do it. "Operations coordinator with four years scheduling 60+ weekly freight runs, applying for the logistics manager role at Hartley & Co" tells a recruiter more than a paragraph of adjectives. Skip objective statements about what you hope to gain; the summary describes what the employer gets.
Most recent job first. Each entry gets the job title, employer, dates, and one to four bullets describing what you did, at what scale, with what result. Ten-year-old roles shrink to a line or two; the recent, relevant ones carry the detail. If you held several roles at one employer, stack the titles under one company heading so the promotion is visible at a glance — that guide shows both formats. If you have no paid work history yet, the section still exists: it fills with volunteering, projects, and coursework, and the first-job CV guide shows exactly how.
Degree, institution, and dates, most recent first. Grades go in when they help and stay off when they don't. If you graduated with distinction, Latin honors like summa cum laude belong on the same line as the degree, lowercase. Students and recent graduates move this whole section above work experience, since the degree is the strongest verified credential they have — the student CV examples show that order on three real one-page CVs. Researchers heading for PhD or faculty posts need the longer academic CV, which adds publications and references.
List the skills the job ad asks for that you genuinely have, in the ad's own wording, with hard skills (software, languages, licences) separated from soft ones. No skill bars, no star ratings — screening software reads "four stars in communication" as nothing. Certifications, languages, and a short interests list follow as their own sections when they add signal; Purdue's OWL resume resources cover the conventions for these smaller sections if you want a second reference.
The unit of a resume is the bullet, and the pattern for a good one is stable: an action verb, the thing you did, and a number that proves scale or outcome. Harvard's career center teaches the same accomplishment-first approach — describe results you can point to rather than duties you were assigned. Two rewrites show the difference:
Before: Responsible for the company's social media accounts. After: Planned and published four posts a week across two channels for an 11k-follower account; engagement rose by a third in six months.
Before: Handled customer service in a busy store. After: Served 80+ customers per weekend shift, resolved refunds and complaints at the till, and trained two new starters.
Both "after" versions describe the same work as the "before" versions. They read better because they contain facts a recruiter could verify in an interview, and that's the standard for every line: if a bullet has no number, ask how many, how often, or how big, and one usually appears. If it genuinely has no checkable content, cut it.
One base resume, adjusted for each job ad, beats one generic resume sent everywhere. The adjustment is smaller than people fear: reread the ad, swap your summary's target role, and align the skills section with the ad's exact vocabulary — if they say "stakeholder management" and you wrote "client communication," use their term where it's honest. This takes about ten minutes per application and directly improves both the keyword match and the human first impression.
AI can speed this up, with one caution: language models fill gaps with plausible inventions, and a fabricated metric follows you into the interview. Our guide to ChatGPT resume writing lays out a workflow that keeps the editing help and drops the fiction. Prezumi's own AI editor takes a different route — it can only reword the profile data you entered, so there's nothing for it to invent.
Before a recruiter sees your resume, a parser usually extracts its text and maps it into fields, and decorative formatting is what breaks that step. The rules that keep extraction clean: a single-column layout, standard headings ("Work Experience", "Education", "Skills"), no tables or text boxes around your history, and a PDF export with real text rather than images of text.
This is testable rather than a matter of faith. Prezumi's free ATS resume checker runs real PDF text extraction on your file and shows you what a parser pulls out, in what order. If you'd rather start from a layout that already passes, the ATS-friendly templates are extraction-tested single-column designs.
Prezumi holds your work history as one profile and renders it into any of 55+ templates, so the writing work above happens once. Import an existing CV to skip retyping, pick an extraction-tested template, and export an ATS-friendly PDF for free. When a specific application needs different wording, the AI editor rewrites your real entries while you watch, and the ATS checker confirms the file parses before you send it. The same profile also renders as a personal website — if your field expects one, the portfolio website guide covers what to put on it.
For most applicants with work history, the order is: header with name and contact details, a two-to-three-sentence summary, work experience in reverse-chronological order, education, then skills, with optional sections such as certifications, languages, and projects after that. Work experience sits high because it is where recruiters spend most of a first scan. Two groups should change the order. Students and first-time applicants move education above experience, because a degree in progress or just completed is their strongest verified credential and their experience section is usually built from projects and volunteering rather than jobs. Academics applying for research posts use a longer document entirely: an academic CV leads with education, then publications, research experience, teaching, and grants, with references listed near the end. Whatever order you choose, keep the section headings standard — "Work Experience", "Education", "Skills" — because applicant tracking systems map your text into fields by recognising exactly those headings.
One page for most applicants, including students, career changers, and anyone with less than roughly ten years of experience. Recruiters spend only a few seconds on a first scan, and a tight single page concentrates those seconds on your strongest material. Two pages are justified when ten or more years of directly relevant experience genuinely fill them — a senior engineer with four substantial roles should use the space rather than shrink the font until the page technically fits. What never works is padding: a second page holding one orphaned section reads as a formatting mistake, so either cut back to one page or fill the second one properly. Academic CVs are the exception to every cap, because hiring committees expect the complete record of publications, teaching, grants, and conference talks, and those documents routinely run to many pages. If you need to trim, cut the oldest roles first; work from more than fifteen years ago rarely changes a hiring decision.
In the US and Canada, "resume" means the one-page job application document and "CV" means the long academic record. In the UK and most of Europe, "CV" is simply the word for the same one-page document. So for normal job applications the two names describe the same thing, and only academic applications need the long-form version.
Yes, two or three sentences. A first scan lasts seconds, and the summary is the part you control most: name the role you're applying for, your strongest credential, and one specific achievement with a number in it. Skip objective statements about your own goals — recruiters read the summary to learn what they get, and generic openers read as filler.
PDF, unless the job posting asks for something else. A PDF preserves your layout on every screen, and modern applicant tracking systems parse text-based PDFs reliably. The one requirement is that the PDF contains real extractable text — a scanned image or a design-tool export that flattens text to graphics will parse as empty. Run the file through a text-extraction checker to confirm.
AI Resume Builder and Portfolio Maker
Build a resume and portfolio free in under 5 minutes: upload your CV, choose a design, then refine every detail — unlimited edits, plus free AI credits for the AI editor. Download your resume PDF free — no hidden paywall.