{"course":{"slug":"ai-careers","title":"Working in AI","url":"https://addaly.com/learn/ai-careers","level":"beginner","track":"work"},"lesson":{"slug":"the-cv-that-survives","title":"A CV that survives a thirty-second read","url":"https://addaly.com/learn/ai-careers/the-cv-that-survives","position":34,"of":85,"minutes":8,"answer":"A CV is scanned for scope, outcome and a number; formatting exists only to make those findable fast.","keyIdea":"A CV is scanned for scope, outcome and a number; formatting exists only to make those findable fast."},"body":{"format":"markdown","text":"## What actually happens to it\n\nYour CV gets well under a minute on the first pass, often nearer twenty seconds. In that time someone is answering four questions: what is this person now, have they done anything close to this, is there a number anywhere, and can I see something they made.\n\nEverything below follows from that.\n\n## The structure that works\n\n1. **Name, one line of positioning, and links.** The positioning line is not a summary paragraph. It is eight words: *Analyst: hospital operations data, SQL and Power BI.* The links are your deployed project first, GitHub second, LinkedIn third.\n2. **Whichever is stronger: experience or selected work.** If you have no relevant jobs, put projects first and call the section **Selected work**, not \"Personal projects\": the second phrasing invites the reader to discount it.\n3. **Skills**, grouped and honest. Three groups, not a wall of forty words.\n4. **Education last**, unless it is your strongest asset.\n\nOne page under about eight years of experience. Two only if the second page earns itself.\n\n## Bullets: what, how, result, number\n\nThis is the whole craft, and it is a rewriting exercise.\n\nBefore: *Responsible for data cleaning and analysis for the sales team.*\n\nAfter: *Rebuilt the weekly sales report from three exported spreadsheets into one scheduled SQL job: cut preparation from four hours to fifteen minutes, and removed a double-counting bug that had inflated Q2 revenue by about eight per cent.*\n\nThe second is not a better-written version of the first. It contains information the first does not: what the situation was, what you did, and what changed. Every bullet should survive the question *so what happened?*\n\nIf you have no numbers, use scale instead: how many people, how many rows, how often it runs, how long it took. *Used by six dispatchers, weekly, since March* is a number.\n\n## Applicant tracking systems, without the folklore\n\nAn enormous amount of nonsense is sold around this. The reality is dull. Tracking systems are text extractors that pull your document into fields. What actually breaks them:\n\n- Text inside an image, including a header rendered as a graphic. There is nothing to extract.\n- Multi-column layouts, where extraction interleaves the columns into unreadable order.\n- A PDF exported from a design tool with no text layer.\n- Section headings the parser does not recognise: \"My Journey\" instead of \"Experience\".\n- Sometimes tables, depending on the parser.\n\nWhat does not break them: a normal PDF exported from a word processor, in one column, with standard headings.\n\nSo \"ATS-optimised templates\" sold for a fee are mostly charging you to avoid a problem you would not have had. Free path: Google Docs, LibreOffice Writer, or a plain LaTeX template. Paid: Canva Pro and subscription resume builders, and a Canva export can parse *worse* than a Word document, because of how the text layer is produced.\n\n## Keywords\n\nMatch the advert's vocabulary where it is true of you. If the posting says \"Power BI\" and you use Power BI, write \"Power BI\": not \"BI tooling\", not \"data visualisation platforms\". A human is scanning for the same word the requisition used, and so is the parser.\n\nThat is matching. Stuffing (listing tools you have not used, or a white-text keyword block) is checkable in about one interview question, and it ends the process.\n\n## With no experience at all\n\nLead with **Selected work**. Three projects, three lines each: what it does and for whom, the hard part, the number. That is the same structure as the case study page, compressed.\n\nInclude non-technical work that shows you can be relied on. Two years in a call centre is evidence of dealing with people under pressure and following process, and a hiring manager for an analyst role reads it as such. Do not hide it; frame it.\n\n## The one-line tailoring rule\n\nDo not rewrite your CV for every application; you will stop applying. Change two things: the positioning line at the top, and the order of your bullets so the most relevant sits first. That is fifteen minutes, not two hours, and it captures most of the benefit.\n\n## The limitation worth knowing\n\nA good CV changes your outcome at the margin: perhaps a few percentage points of response rate. The *route* changes it by an order of magnitude: a referred application is read by a human within a day, and a cold one joins several hundred others. Spend an evening making the CV good, then spend the rest of your evenings on the routes in the other lessons of this module. People polish CVs for months because it is the only part of the search that feels controllable."},"check":{"question":"Kwame has no paid experience in data. His CV opens with a three-sentence career objective, then education, then a section titled \"Personal projects\" listing four projects by name and technology. Which single change is likely to help most?","options":[{"text":"Convert the CV to a designed template with a sidebar and icons, so it stands out visually.","correct":false,"tempting":"Optimises for attention rather than evidence, and multi-column designs are exactly what text extraction handles worst."},{"text":"Add every relevant tool from the job advert to a skills section to improve keyword matching.","correct":false,"tempting":"Chases the parser instead of the reader, and adds tools he cannot defend in an interview."},{"text":"Replace the objective with an eight-word positioning line and links, move the work to the top under a heading like \"Selected work\", and give each project three lines saying who uses it, what was hard, and one number.","correct":true},{"text":"Remove the projects section entirely, since unpaid work is discounted by employers anyway.","correct":false,"tempting":"Deletes the only evidence he has, on a belief about employers that is true of some large filters and false of the people who would actually hire him."}]},"prev":{"slug":"the-numbers-of-a-search","title":"Running the search as a funnel","url":"https://addaly.com/learn/ai-careers/the-numbers-of-a-search"},"next":{"slug":"the-profile-people-open","title":"The profile a stranger checks in ninety seconds","url":"https://addaly.com/learn/ai-careers/the-profile-people-open"},"authorship":"Founder and series editor: Dr Mustafa Mun. Written by AI models to a written standard. Exam keys checked blind by a second model: 3,665 of 3,666 agreed. Academic review invited.","review":{"status":"Draft, open for academic review"},"cite":"Addaly (2026) 'A CV that survives a thirty-second read', in Working in AI. Edited by M. Mun. Addaly. Available at: https://addaly.com/learn/ai-careers/the-cv-that-survives (Accessed: 13 October 2026).","licence":{"name":"CC BY-NC-SA 4.0","url":"https://creativecommons.org/licenses/by-nc-sa/4.0/"},"reader":{"schema":1,"title":"A CV that survives a thirty-second read","courseTitle":"Working in AI","module":{"index":4,"slug":"getting-seen","title":"Getting seen: the routes that put a person in front of your work"},"answer":null,"keyIdea":"A CV is scanned for scope, outcome and a number; formatting exists only to make those findable fast.","hero":{"kind":"plate","image":null,"alt":null,"caption":null},"blocks":[{"t":"h2","id":"what-actually-happens-to-it","runs":[{"text":"What actually happens to it"}]},{"t":"p","runs":[{"text":"Your CV gets well under a minute on the first pass, often nearer twenty seconds. In that time someone is answering four questions: what is this person now, have they done anything close to this, is there a number anywhere, and can I see something they made."}]},{"t":"p","runs":[{"text":"Everything below follows from that."}]},{"t":"h2","id":"the-structure-that-works","runs":[{"text":"The structure that works"}]},{"t":"ol","start":1,"items":[[{"text":"Name, one line of positioning, and links.","b":true},{"text":" The positioning line is not a summary paragraph. It is eight words: "},{"text":"Analyst: hospital operations data, SQL and Power BI.","i":true},{"text":" The links are your deployed project first, GitHub second, LinkedIn third."}],[{"text":"Whichever is stronger: experience or selected work.","b":true},{"text":" If you have no relevant jobs, put projects first and call the section "},{"text":"Selected work","b":true},{"text":", not “Personal projects”: the second phrasing invites the reader to discount it."}],[{"text":"Skills","b":true},{"text":", grouped and honest. Three groups, not a wall of forty words."}],[{"text":"Education last","b":true},{"text":", unless it is your strongest asset."}]]},{"t":"p","runs":[{"text":"One page under about eight years of experience. Two only if the second page earns itself."}]},{"t":"h2","id":"bullets-what-how-result-number","runs":[{"text":"Bullets: what, how, result, number"}]},{"t":"p","runs":[{"text":"This is the whole craft, and it is a rewriting exercise."}]},{"t":"p","runs":[{"text":"Before: "},{"text":"Responsible for data cleaning and analysis for the sales team.","i":true}]},{"t":"p","runs":[{"text":"After: "},{"text":"Rebuilt the weekly sales report from three exported spreadsheets into one scheduled SQL job: cut preparation from four hours to fifteen minutes, and removed a double-counting bug that had inflated Q2 revenue by about eight per cent.","i":true}]},{"t":"p","runs":[{"text":"The second is not a better-written version of the first. It contains information the first does not: what the situation was, what you did, and what changed. Every bullet should survive the question "},{"text":"so what happened?","i":true}]},{"t":"p","runs":[{"text":"If you have no numbers, use scale instead: how many people, how many rows, how often it runs, how long it took. "},{"text":"Used by six dispatchers, weekly, since March","i":true},{"text":" is a number."}]},{"t":"h2","id":"applicant-tracking-systems-without-the-folklore","runs":[{"text":"Applicant tracking systems, without the folklore"}]},{"t":"p","runs":[{"text":"An enormous amount of nonsense is sold around this. The reality is dull. Tracking systems are text extractors that pull your document into fields. What actually breaks them:"}]},{"t":"ul","items":[[{"text":"Text inside an image, including a header rendered as a graphic. There is nothing to extract."}],[{"text":"Multi-column layouts, where extraction interleaves the columns into unreadable order."}],[{"text":"A PDF exported from a design tool with no text layer."}],[{"text":"Section headings the parser does not recognise: “My Journey” instead of “Experience”."}],[{"text":"Sometimes tables, depending on the parser."}]]},{"t":"p","runs":[{"text":"What does not break them: a normal PDF exported from a word processor, in one column, with standard headings."}]},{"t":"p","runs":[{"text":"So “ATS-optimised templates” sold for a fee are mostly charging you to avoid a problem you would not have had. Free path: Google Docs, LibreOffice Writer, or a plain LaTeX template. Paid: Canva Pro and subscription resume builders, and a Canva export can parse "},{"text":"worse","i":true},{"text":" than a Word document, because of how the text layer is produced."}]},{"t":"h2","id":"keywords","runs":[{"text":"Keywords"}]},{"t":"p","runs":[{"text":"Match the advert’s vocabulary where it is true of you. If the posting says “Power BI” and you use Power BI, write “Power BI”: not “BI tooling”, not “data visualisation platforms”. A human is scanning for the same word the requisition used, and so is the parser."}]},{"t":"p","runs":[{"text":"That is matching. Stuffing (listing tools you have not used, or a white-text keyword block) is checkable in about one interview question, and it ends the process."}]},{"t":"h2","id":"with-no-experience-at-all","runs":[{"text":"With no experience at all"}]},{"t":"p","runs":[{"text":"Lead with "},{"text":"Selected work","b":true},{"text":". Three projects, three lines each: what it does and for whom, the hard part, the number. That is the same structure as the case study page, compressed."}]},{"t":"p","runs":[{"text":"Include non-technical work that shows you can be relied on. Two years in a call centre is evidence of dealing with people under pressure and following process, and a hiring manager for an analyst role reads it as such. Do not hide it; frame it."}]},{"t":"h2","id":"the-one-line-tailoring-rule","runs":[{"text":"The one-line tailoring rule"}]},{"t":"p","runs":[{"text":"Do not rewrite your CV for every application; you will stop applying. Change two things: the positioning line at the top, and the order of your bullets so the most relevant sits first. That is fifteen minutes, not two hours, and it captures most of the benefit."}]},{"t":"h2","id":"the-limitation-worth-knowing","runs":[{"text":"The limitation worth knowing"}],"role":"limit"},{"t":"p","runs":[{"text":"A good CV changes your outcome at the margin: perhaps a few percentage points of response rate. The "},{"text":"route","i":true},{"text":" changes it by an order of magnitude: a referred application is read by a human within a day, and a cold one joins several hundred others. Spend an evening making the CV good, then spend the rest of your evenings on the routes in the other lessons of this module. People polish CVs for months because it is the only part of the search that feels controllable."}]}],"check":{"question":[{"text":"Kwame has no paid experience in data. His CV opens with a three-sentence career objective, then education, then a section titled “Personal projects” listing four projects by name and technology. Which single change is likely to help most?"}],"options":[{"text":[{"text":"Convert the CV to a designed template with a sidebar and icons, so it stands out visually."}],"correct":false,"tempting":[{"text":"Optimises for attention rather than evidence, and multi-column designs are exactly what text extraction handles worst."}]},{"text":[{"text":"Add every relevant tool from the job advert to a skills section to improve keyword matching."}],"correct":false,"tempting":[{"text":"Chases the parser instead of the reader, and adds tools he cannot defend in an interview."}]},{"text":[{"text":"Replace the objective with an eight-word positioning line and links, move the work to the top under a heading like “Selected work”, and give each project three lines saying who uses it, what was hard, and one number."}],"correct":true,"tempting":null},{"text":[{"text":"Remove the projects section entirely, since unpaid work is discounted by employers anyway."}],"correct":false,"tempting":[{"text":"Deletes the only evidence he has, on a belief about employers that is true of some large filters and false of the people who would actually hire him."}]}],"order":[3,0,2,1],"why":null},"prev":{"slug":"the-numbers-of-a-search","title":"Running the search as a funnel","minutes":9},"next":{"slug":"the-profile-people-open","title":"The profile a stranger checks in ninety seconds","minutes":8}}}