CV Parsing
CV parsing is the automatic extraction of information from a CV or resume, such as contact details, work history, education and skills, into structured fields in a database or applicant tracking system.
Definition
CV parsing turns a document that is written for humans, in many layouts and formats, into data a system can search and sort. The software reads the file, identifies sections and extracts items such as job titles, employers, dates, degrees, languages and skills, then fills the matching fields in a candidate profile.
Its quality is measured by how often the extracted fields are correct and complete, which depends on the file format, the layout, the language and unusual structures such as columns, tables or graphics. A sensible practice is to test the parser on a sample of real CVs from your own candidates and to let candidates correct their profile.
Parsing saves time on data entry and makes search possible, but it has limits. Errors can send a candidate to the wrong place, for instance by misreading dates, and any ranking built on parsed data inherits those errors and the biases in the criteria. Parsing is an input for human review, not a judgement on the candidate. Personal data in CVs must be handled according to the data protection rules that apply to you.
Key points
- It extracts structured data such as roles, dates, education and skills from CV files.
- Accuracy varies with layout, language and file format: test it on your own CVs.
- It saves manual data entry and makes candidate databases searchable.
- Errors carry through to any search or ranking, so keep people reviewing the results.
- CVs contain personal data: apply the data protection rules that apply to you.
Example
A recruiter receives 200 CVs for an accountant role. After parsing, the tool lists each candidate's last job title, years of experience and software mentioned. A quick check on 20 files shows that CVs with two columns often have their dates mixed up, so the team asks for a plain format and checks those profiles by hand.
Frequently asked questions
What formats can a CV parser read?
Most parsers handle common formats such as PDF and Word documents, but quality varies. Scanned documents, images and heavily designed layouts are the hardest to read correctly.
Is CV parsing the same as AI screening?
No. Parsing only extracts and organises information. Screening is the evaluation of candidates against criteria, which can use parsed data but should remain a human decision.
Can parsing introduce bias?
Parsing itself does not judge, but errors in reading some names, languages or formats can disadvantage some candidates. Test it on varied CVs and review borderline cases yourself.
Related terms
AI conversations, human decisions
Passisto holds structured voice conversations with candidates, employees, clients and students, and turns them into reports your team can act on.