A resume parser that returns a working profile, not a blob of text

Aptora reads a CV from a PDF, a screenshot or a phone photo and turns it into a structured profile: positions with dates, education, projects, contact details, and the capabilities found in the prose. You get editable fields in a defined schema — not a wall of extracted text you still have to sort yourself.

Profile in AptoraCV format (excerpt)

{
  "meta": { "format": "AptoraCV@1.0.0" },
  "name": "Jana Weber",
  "info": { "label": "Senior Platform Engineer" },
  "location": { "city": "München", "country": "DE" },
  "employment": { "history": [
    { "position": "Platform Engineer",
      "employer": "Contoso GmbH",
      "start": "2021-03" }
  ] },
  "skills": { "list": [
    { "name": "Kubernetes" },
    { "name": "Terraform" },
    { "name": "Go" }
  ] }
}

Every field lands where it belongs

The parser writes into AptoraCV, a published schema with dedicated sections for employment, education, projects, languages, references and more. A date is a date and a position is a position — so everything downstream, from matching to PDF export, can rely on the structure instead of guessing at strings.

Capabilities are detected, not just copied

Skills are read out of the running text, spelling variants are reconciled, and related capabilities are linked. Someone who runs Kubernetes clusters does not need to have typed the word "Docker" for Aptora to understand what they can do.

Bad scans are a normal case, not an error

Screenshots and phone photos are read the same way as clean PDFs. When a document is genuinely unreadable, Aptora says so instead of inventing content — and everything it did extract is shown for review before a profile is created.

From parsed to useful in one step

The parsed result is a live profile: render it in a theme, export it as a PDF, publish it at a share link, or score it against real job listings. Parsing is the entrance, not the product.

Frequently asked questions

Which file formats can the resume parser read?

PDF, PNG and JPEG — including screenshots and phone photos of a printed CV. The document is read as a whole, so multi-column layouts and design-heavy templates work too.

What structure does the parsed result have?

A JSON document in the AptoraCV schema: personal details, employment history with dates, education, projects, skills, languages, references and further sections. The full field list is public on the schema page.

Can I correct what the parser got wrong?

Yes. The result opens in a review step before anything is saved, and every field stays editable afterwards. Parsing mistakes never lock you in.

How do I use Aptora with my AI assistant?

Connect your AI assistant to Aptora over MCP. It can validate, create and edit structured profile data. Upload files in the Aptora app; if your AI client can read documents, it can also prepare a profile draft from them. The MCP server itself does not accept PDF or image files.

Get started