What we do — and what we don't.
We built getshortlisted with a clear opinion of what a CV tool should be — and what it shouldn't try to be. Here's the long version of what's on the homepage.
What it does
- Turns your work history into a clean schema.
- Your CV is built from a schema, not a Word template. Each role, each project, each skill is its own field — so the form catches gaps, the export stays consistent, and you can rearrange without retyping. The structure does the boring work; you keep the writing.
- Edits your draft — doesn't write it for you.
- The AI is a copyeditor, not a ghostwriter. It tightens what you've drafted, suggests sharper verbs, and flags lines that read like everyone else's. It doesn't invent achievements, it doesn't generate filler, and it never replaces the page with a synthesised version of you. If a recruiter could tell it was AI, that's exactly the failure mode we're avoiding.
- Hosted in Switzerland — GDPR-equivalent.
- Your drafts, your exports, the AI calls themselves — all run on infrastructure hosted in Switzerland by AlexHost. Switzerland has GDPR adequacy under EU Commission Decision 2000/518/EC, so the protections apply by default. Your CV is your document; we don't sell, share, or train on it.
What it doesn't do
- No auto-apply — you stay in control of every submission.
- We don't submit applications on your behalf. The market has tried this — the products that survived are rated 2 out of 5, and the ones that didn't shut down. Recruiters add traps for AI-submitted applications, ATSes flag identical AI letters, and platforms restrict the accounts that automate. Beyond the user impact, the EU AI Act treats employment-decision automation as high-risk under Article 14 (enforceable August 2026), and the Product Liability Directive makes any harm from a hallucinated submission strictly liable on us. None of those costs are worth a feature that gets your application rejected before a human reads it.
- No reading your inbox — we don't ask for that access.
- We don't ask for email access, we don't track replies from recruiters, and we don't surface "insights" from your messages. The OAuth scope it would take is one of the most invasive permissions a product can request — and it would turn us into a data controller for content that has nothing to do with your CV. The cost is wrong, the boundary is wrong, we don't cross it.
- No scraping job sites — paste the description in instead.
- We don't pull listings from LinkedIn, Indeed, or any other platform without their permission. Their terms of service are binding contracts under EU law, and scraping personal data published on those platforms makes us a joint controller with the platform under GDPR — a position we're not willing to take. If you want to tailor your CV to a specific role, paste the description in. That keeps the relationship with the platform clean and yours.
- No score out of 100 — it would optimise the wrong thing.
- An algorithmic CV score is a number designed to make you anxious enough to keep editing. The CV doesn't have an objective score — recruiters don't read it that way, and ATSes don't either. Putting a 73/100 next to your draft would be the wrong incentive: it would optimise for the score instead of the role. We give you concrete, line-level feedback. We don't give you a leaderboard.
These are not roadmap items pending revisit — they are product axioms.