Automatically extract CV data with AI

Dorothée and her HR team read your PDF and DOCX CVs and turn them into structured candidate records: skills, years of experience, education, languages, mobility. No more copy-pasting into your ATS.

Extract my CVs See an extracted record

Manually parsing 50 CVs takes half a day. Dorothée does it in 5 minutes, with a JSON or CSV output directly importable into your ATS (Recruitee, Greenhouse, Workable, Teamtailor). She handles multi-language and recognizes sector-specific certifications.

A recruiter comparing several candidate records on a screen.

Why manual CV parsing is broken

CVs arrive in 15 different formats: native PDF, scanned PDF, DOCX with macros, LinkedIn exports, Pages files, sometimes even images. Recruiters lose 6 minutes on average per CV extracting useful data: name, contact, experiences with calculated durations, explicit and implicit skills, certifications, geographic mobility. On 100 applications, that's a full day lost before the first shortlist. Worse, cognitive bias creeps in: you remember what's at the top, you miss what's at the bottom. A properly configured AI parser does the work in 30 seconds per CV, with no cognitive bias and a standardized output format.

How Dorothée processes your CVs

Dorothée is more than blind OCR. She runs an extraction chain that understands the semantics of a CV.

1
OCR if needed: Dorothée detects whether the CV is a native PDF (text) or a scan (image) and applies OCR only when useful.
2
Structured extraction: name, contact, experiences with automatic duration calculation, education, certifications, languages with level, mobility.
3
Normalization: harmonization of job titles (Senior Dev / Lead Dev / Tech Lead → comparable family), recognition of sector certifications.
4
Optional GDPR anonymization: Rex masks name, photo, age, address and nationality before scoring if you enable anti-bias mode.
5
Export: JSON for API, CSV for Excel/ATS import, direct push via webhook to supported ATSes.

5-step pipeline: OCR, extraction, normalization, anonymization, export. — The full CV extraction flow in 30 seconds per CV.

What you get

Structured candidate record

All standard HR fields, ready to integrate with your ATS

CSV or JSON export

For Excel, Google Sheets, or custom API import

Direct ATS push

Recruitee, Greenhouse, Workable, Teamtailor via webhook

Multi-language

Recognition of CVs in French, English, German, Spanish, Portuguese, Chinese

Job-fit scoring

If you provide the job description, Dorothée scores each CV 0-100

Inconsistency detection

Career gaps, conflicting dates, unverifiable credentials

Mockup of a structured candidate record with skills, experience, education. — The standardized candidate record ready to push to your ATS.

Dorothée's HR team

Dorothée

HR Director & Tutor — leads CV extraction and onboarding

Platform Guide

Interactive onboarding

Enterprise Onboarding

4-phase plan

AI Trainer

3-level training

Rex

GDPR & anonymization

Sage

Sales Compliance GDPR

Pixel

Data Anonymization

Included from Free — 100 CVs/month

Unlimited volume from Startup at €69/month, ATS integrations included.

See all plans

Frequently asked questions

Does Dorothée work on scanned CVs (PDF image)?

Yes — OCR built in for scanned PDFs and images. Extraction quality close to 95% on clean scans, 80% on degraded scans.

Is she GDPR-compliant for processing candidate data?

Yes — European hosting, signable DPA, configurable retention period, deletion on request, candidate access right. Compliant with GDPR Article 22 on automated decisions: Dorothée extracts and scores, but the final hiring decision stays human.

Can she score candidates against a job description?

Yes — provide the job description and Dorothée weighs the required skills, computes a 0-100 score and explains the gaps. Scoring is transparent and auditable, not a black box.

Which languages are supported?

6 native languages: French, English, German, Spanish, Brazilian Portuguese, Chinese. For other languages, Dorothée extracts structured fields but with lower precision on sector skills.

How do you avoid discriminatory bias in extraction?

Dorothée can be configured to automatically anonymize name, photo, age, address and nationality before scoring. This produces an objectified first shortlist; the recruiter reintroduces identity afterward.

First extraction in 30 seconds

100 CVs free per month, no credit card.

Try Dorothée

Other tasks you might like

Task

Automate cold email with an AI agent

Mia drafts, personalizes and follows up your cold emails for you.

Task

Generate SEO articles with an AI agent

Lili and her marketing team generate full SEO articles: keyword research, H1-H3 structure, optimized meta, automatic internal linking, JSON-LD.

Task

Automated cybersecurity audit for SMBs and startups

Guard and his team of 27 specialists scan your infrastructure, flag OWASP vulnerabilities, check your NIS2 and GDPR compliance, and ship an actionable report your engineers can run with.