75% of CVs are filtered out before a human reads them. OracleCV scores yours against the same checks 14 real ATS parsers run, rewrites weak bullets with AI, and ships in any format your next employer wants.
The four sections below cover how OracleCV turns a draft into a parser-friendly, AI-tuned CV. Why most CVs get filtered out, the editor plus live ATS score, the template gallery, and the four-step flow from upload to finished file.
ATS parsers rank by literal keyword overlap. "Led roadmap" and "owned roadmap" are different to a machine. We extract the job description's vocabulary and tell you what's missing.
"Responsible for", "helped with", "worked on". These tell a recruiter you were in the room, not what you did. Our AI rewrites them with strong, specific verbs.
A bullet without a metric is a claim. Adding scale ("8 engineers"), outcome ("+22% conversion"), or duration ("over 14 months") turns it into evidence.
Multi-column tables, text in images, fancy headers. Many ATS systems read them as jumbled noise. Our templates are designed for both audiences.
"My Journey" reads beautifully, and parses as nothing. ATS look for "Experience", "Education", "Skills" by exact match. Stay standard, lead with substance.
A 4-page CV reads as unfocused. A half-page CV at 8 years' experience reads as thin. We benchmark against the role and tell you what to trim or expand.
You don't design anything. You tell the agent in plain English (“more editorial,” “serif headlines, tighter density,” “a monospace data block on the right”) and a beautiful, parser-friendly template is generated for you. Every edit re-runs through real ATS parsers, so the score stays high while the design becomes yours.
Every template uses embedded fonts and a single-column layout. Switch between them with one click. Your content moves with you.
No accounts, no setup. Bring a CV, leave with a version tuned to the job, and a score that says it'll get through.
PDF, DOCX, or plain text. We read it the way real ATS parsers do.
Start →We pull the keywords and weights the parser will look for.
Continue →Twenty frames, all ATS-friendly. Switch any time. Content moves with you.
Continue →AI suggests stronger bullets in your voice. Accept, tweak, or ignore. Your score updates live.
Finish →Once your CV is sharp, OracleCV becomes a job radar. Your profile is broadcast across LinkedIn, Glassdoor, Indeed and 15 more boards in parallel via Exa, Brave, and Tavily. Claude reads each posting, ranks the matches, and tailors a copy of your CV for the role you pick.
Drop in your CV, AI sharpens it for the ATS, the score climbs, then your profile is broadcast across every major job board. The strongest match comes back ranked and ready to apply.
Click Find Jobs on any saved CV and OracleCV fans the query out across Exa, Brave, and Tavily in parallel. Hits get merged by URL. Anything multiple engines return rises to the top. Claude reads each posting and synthesises a per-role match score, a why-match outline, and a list of required skills.
We know the dream job sometimes takes a while to find. So every plan scales down in price the longer you stay. Quarterly is cheaper than monthly per month, yearly is cheaper still. Keep the radar on for as long as you need; the meter slows down the deeper you go.
Billed monthly · €25.00 every month
Billed every 3 months · €60.75 per quarter
Billed yearly · €180.00 per year
Refund any first payment within 14 days. No forms, no retention call.
Refund policy →One click in the Stripe portal. Cancellation kicks in at the end of the current period.
We never see your card. Stripe handles every charge under PCI-DSS Level 1.
No third-party ads, no data brokers. GDPR rights honored at /privacy.
Privacy policy →Yes. The core flow is free: upload an existing CV, score it against the same parsers an ATS uses, match it to a job description, and export an ATS-friendly PDF. There is no paywall on the score or on basic exports.
We run six independent sub-scores: parseability, skill match, experience match, education match, outcome density, and contact-block hygiene. Each sub-score returns the specific evidence we used, so you can see which line dragged the score down, not just a single opaque number.
Yes. Paste the JD or upload the posting and we compute a per-section match: which required skills are present, which preferred skills are missing, whether your experience meets the minimum years, and whether your degree level satisfies the role. The score updates live as you edit.
The templates are constructed so the underlying text remains the single source of truth: no images, no two-column tricks that confuse parsers, embedded standard fonts, and selectable PDF text. Every render is re-parsed and re-scored so a template choice that hurts parseability is flagged before you download.
PDF (recruiter-ready, ATS-safe), DOCX (Word), TXT (plain text for legacy parsers), Markdown (for portfolios), and JSON (a typed schema you can pipe into other tools).
Both, scoped by you. You can highlight a single bullet and ask for a stronger version, or rewrite a full experience block to target a specific job description. The AI uses Claude under the hood and only changes the lines you point at.
Your CV is stored in your private workspace, never used to train models, and never shown publicly. The score itself is computed deterministically on our parsers; only the AI rewrite step calls an external LLM, and only on the snippet you select.
About 90 seconds for an upload and score, and 5 to 10 minutes for a full pass: upload, score, paste a JD, accept AI rewrites on a few bullets, swap to a stronger template, and export.
Cancel anytime, in two clicks. No retention emails, no exit survey. Come back for the next move.
No. Your CV is encrypted at rest and never used to train models. You can delete everything in one click.
Yes. Upgrade or downgrade at the next billing cycle. The unused time is prorated.
Upload your CV. See your score. Fix it in 10 minutes.
