Ashby CV tips: evidence the AI can cite (2026)
Ashby is the ATS behind many of tech's best-known hiring pipelines, and its AI review works on criteria and evidence, not keywords.
How Ashby handles your CV
Ashby is an all-in-one recruiting platform (ATS, CRM, scheduling, structured interviews, analytics) adopted widely by technology companies and high-growth startups. Applying is a short form with no account creation, and your CV is parsed to populate the candidate record.
What sets Ashby apart is its AI-assisted application review. Instead of parsing your CV into rigid fields and matching keywords, the AI reads the whole document against the criteria the hiring team defined for the role and reports which criteria you meet, citing the specific evidence in your CV for each. Recruiters review those cited findings; the AI does not reject anyone.
This inverts the old ATS logic. A keyword with no supporting sentence produces no citation and therefore nothing. A plainly written achievement that demonstrates a criterion is exactly what the system is built to find, wherever it sits in the document.
What to do: the Ashby checklist
- Give each criterion in the advert one quotable line: a self-contained sentence pairing the skill with a concrete outcome, because Ashby's AI review cites evidence and needs a sentence it can quote.
- Name the subject explicitly inside the sentence ("Reduced AWS spend 30% by rearchitecting the data pipeline" rather than "reduced costs significantly"), so a citation stands alone when a recruiter reads it out of context.
- Do not keyword-stuff: unsupported terms produce no citation in Ashby, and the recruiter sees the cited evidence, not a keyword count.
- Answer the form's questions properly; they sit alongside the AI review in what the team sees.
- Keep the layout simple with a real text layer; the AI reads the extracted text, and a broken extraction has nothing to cite.
- Assume a structured process end to end: Ashby teams typically run scorecard-based interviews, so the claims in your CV will be tested against the same criteria later.
Myth vs reality
The myth
AI-native ATSs like Ashby are beaten by heavier keyword optimisation.
The reality
The opposite. Ashby's review is criterion-and-evidence based: it reads your CV like a careful human and cites the sentences that demonstrate each requirement. A keyword without a supporting sentence yields nothing citable, while a plainly written, quantified achievement is surfaced automatically. Optimise for evidence per requirement, not term frequency.
How Vouch prepares your CV for Ashby
Ashby rewards what Vouch is built to produce: claims backed by real stories. Maya's interview extracts the concrete evidence behind your skills, and Vouch writes achievement lines that pair skill with outcome, the exact shape an evidence-citing reviewer, human or AI, can lift and verify. Anything you did not actually say never makes it into the CV.
Every CV Vouch exports is a single-column layout with a real text layer, generated from structured data rather than a screenshot, so any parser in this guide can read it. Start with the free cover letter generator, and if you are converting a US resume for UK employers, see UK CV vs American resume.
Frequently asked questions
Does Ashby auto-reject applications?
No. As of August 2026 Ashby's AI-assisted review reads applications against the role's criteria and reports the evidence it found, but accept and reject decisions belong to the hiring team. The AI's output is a set of cited findings a recruiter checks, which is why evidence-rich sentences matter more than anything else in an Ashby application.
Which companies use Ashby?
Ashby is concentrated in technology: OpenAI, Notion, Ramp, Linear, Deel, Reddit, Zapier and many high-growth startups run their hiring on it as of 2026. If you are applying to a well-known tech company or a venture-backed startup and the application form is short, clean and account-free, there is a fair chance it is Ashby (the URL often contains ashbyhq.com).
How should I format a CV for Ashby?
Simple single-column layout, real text layer, and one self-contained, quantified sentence per skill the advert asks for. Ashby's AI reads the full text against the role's criteria and cites supporting sentences, so structure matters less than sentence quality: each claim should stand alone with the skill, the action and the measurable outcome in it.