Every suggestion in the app comes from a rule in our library, with its sources attached. The library is curated and reviewed by a person; the AI never writes its own citations.
Applying outside the US?
Our defaults follow US resume conventions. Photos, personal details, dates and length differ by country and employer, in the UK, continental Europe, the Baltics and Nordics. Regulated jobs may need licences, right-to-work or clearances stated even when general advice says to leave personal details out.
Example rules
A few rules from the library. The rest appear in the app, with their sources, when they apply to your resume.
R-GAPS-01
Make your experience easy to see: show how long you held each role. Never falsify dates.
In a UK field experiment with 9,022 real vacancies, listing years worked per job instead of exact dates raised callbacks by 8%, and by 15% for applicants with a career gap.
- Applies to
- Tested in the UK; norms differ by country and employer.
- Note
- Some employers and application forms require exact dates; use them when asked.
- Reducing discrimination against job seekers with and without employment gaps (Nature Human Behaviour) (2023)
Preregistered UK field experiment, 9,022 vacancies: listing years worked per job instead of exact dates raised callbacks about 8% (no gap) and 15% (with a gap).
Reviewed 2026-11-10
R-VOICE-01
Prefer light edits that keep your facts, voice and wording; full rewrites only if you ask.
In one study, an AI screener picked the version of a resume that the same AI had rewritten over the human original 68–88% of the time, and some models also favour their own writing over other AIs'. Usually you can't know which AI an employer uses, so by default we don't tune your resume to one model's style: light edits keep your facts specific and readable to any screener. If you do know, for example you're applying to an AI company that likely screens with its own model, your edits can be written with that model.
- Applies to
- All resumes.
- Note
- Ask for a full rewrite any time.
- AI Self-preferencing in Algorithmic Hiring: Empirical Evidence and Insights (2025) · Preprint; extended abstract at AIES 2025
Each AI model, acting as screener, picked the version of a resume that it had rewritten itself over the human original 68–88% of the time, with content quality controlled.
- Making Talk Cheap: Generative AI and Labor Market Signaling (2025) · Preprint
When AI made tailored applications cheap, they stopped signalling ability and hiring became less meritocratic.
Reviewed 2026-11-10
R-LENGTH-01
Choose length by career stage, sector and relevant content. There's no universal one-page rule.
Accounting recruiters ranked two-page resumes above one-page ones for equally qualified candidates, including recruiters who said they preferred one page.
- Applies to
- Depends on seniority, sector and country.
- Note
- Follow any page limit an employer states.
- One Page or Two? A National Study of CPA Recruiters' Preferences for Résumé Length (Journal of Business Communication) (2001)
Accounting recruiters ranked two-page resumes above one-page ones for similarly qualified candidates, even when they said they preferred one page.
Reviewed 2026-11-10
R-CLARITY-01
Fix spelling, grammar and usage while keeping your meaning.
In a field experiment with about 480,000 jobseekers, those given spelling and grammar help were hired 8% more often, at 10% higher wages, with no drop in employer satisfaction.
- Applies to
- Tested on an online labour market; plausible for most written applications.
- Note
- The help fixed errors; it didn't rewrite.
- Algorithmic Writing Assistance on Jobseekers' Resumes Increases Hires (Management Science) (2025)
Field experiment, ~480,000 jobseekers on an online labour market: spelling and grammar help raised hires 8%, at 10% higher wages, with no drop in employer satisfaction.
Reviewed 2026-11-10