Rule-based parsing · 1 credit per CV · 56 fields

Rule-Based CV Parsing for High-Volume Recruitment

Turn information in CVs into structured candidate records with deterministic extraction. Choose an AI mode when a document or the information you need requires more context.

Try it free with 200 credits Processing many CVs? Book a demo

Rule-based parsing · 1 credit per CV · 56 fields

What is rule-based CV parsing?

CV parsing turns a document written for people to read into structured data for recruiting software. The rule-based mode identifies and normalizes information found in a CV to create a candidate record for your ATS, database or internal workflows.

What information can a CV parser structure?

The rule-based mode provides 56 fields for organizing information detected in a document.

Identity and contact details

Name and contact information included in the CV.

Work experience

Roles, employers, dates and detected career details.

Education

Degrees, institutions and dates shown in the document.

Skills

Skills extracted and organized in the candidate profile.

Languages and other data

Other available fields when the CV contains matching information.

56 is the number of fields available, not a guarantee that every CV will fill all of them. Results depend on the information in the document.

From CV to structured candidate record

The engine reads the document, finds supported information and maps it to consistent fields for review and use.

1. Read the document
Extract usable content from the submitted CV.
2. Find the data
Identify contact details, skills, experience and other available information.
3. Normalize
Organize detected values into the matching candidate fields.
4. Use the result
Review the record, then use it in CVReaderPro, an ATS or another workflow.

A rule-based engine that can evolve with feedback

Parsing a CV and improving the engine are separate steps.

While parsing with rules

Extraction uses the engine’s rules and normalization methods. An AI model is not called for every CV processed in this mode.

To improve the engine

Validated manual corrections and AI parsing results can help identify cases to improve and inform changes to the rule-based engine.

Corrections and results provide feedback for review. This does not mean a rule changes automatically after every correction or that an AI model is used during rule-based parsing.

Review a structured output example

The synthetic examples below let you inspect the returned data format. They illustrate the output structure and are not product performance measurements.

Fictional CV

Source document

Claire Martin

Java Developer · Lyon

Experience: Java Developer — NovaTech · since March 2022
Software Developer — Digital Solutions · 2019–2022
Education: MSc Computer Science — University of Lyon
Skills: Java · Spring Boot · SQL · Git · Docker

Structured data preview

Illustrative JSON

{
  "candidate": {
    "firstName": "Claire",
    "lastName": "Martin",
    "jobTitle": "Java Developer",
    "city": "Lyon"
  },
  "experiences": [
    {
      "mission": "Java Developer",
      "company": "NovaTech",
      "startMonth": "03",
      "startYear": "2022",
      "endMonth": "",
      "endYear": "",
      "currentPosition": true
    },
    {
      "mission": "Software Developer",
      "company": "Digital Solutions",
      "startMonth": "09",
      "startYear": "2019",
      "endMonth": "02",
      "endYear": "2022",
      "currentPosition": false
    }
  ],
  "educations": [
    {
      "dateReceived": "2019",
      "title": "MSc Computer Science",
      "school": "University of Lyon",
      "description": ""
    }
  ],
  "skills": [
    {
      "name": "Java",
      "level": ""
    },
    {
      "name": "Spring Boot",
      "level": ""
    },
    {
      "name": "SQL",
      "level": ""
    },
    {
      "name": "Git",
      "level": ""
    },
    {
      "name": "Docker",
      "level": ""
    }
  ]
}

Educational example, not a performance measurement. Real results depend on the document and the information it contains. Recorded product examples follow below.

Explore a synthetic example

Download the CV inputs and captured API outputs. All people, employers and vacancies are fictional. The sample documents are in English; results depend on the input and parser and require human review.

Sample CV (PDF)Rule-Based result (JSON)AI sample CV (PDF)AI 56 result (JSON)

When is rule-based CV parsing a good fit?

This mode fits organizations that need to turn a stream of CVs into consistent candidate data, especially when volume and integration matter.

Recruitment agencies

Structure incoming CVs to populate candidate records without re-entering information already in each document.

IT services and staffing firms

Turn CVs and consultant profiles into structured data before search, matching or client CV preparation.

ATS vendors

Add CV extraction to your application and reserve AI modes for cases that need more context.

HR teams

Bring key CV details into consistent profiles instead of working only with unstructured documents.

Rule-based or AI CV parsing: which mode should you choose?

Choose a processing level based on the document, the context required and the cost per CV.

Mode	Credits / CV	Fields	Extraction type	Best suited when
Rule-based parsing	1	56	Deterministic rules	Standardized extraction is enough or you process CVs at scale.
AI parsing 56	3	56	AI model with more context	The CV requires more contextual interpretation.
AI parsing 96	5	96	More extensive AI interpretation	You need a more detailed candidate record.

Fields populated depend on the information available in each source CV. AI parsing does not need to replace rule-based parsing in every workflow.

Compare API options View the documented JSON schema Estimate your credit needs

Supported formats and parsing limitations

CVReaderPro public demo flows accept formats including:

PDF
DOC
DOCX
TXT

Keep in mind

Information missing from a CV cannot be invented or filled in as if it were present.
The number of populated fields varies from one document to another.
Document structure and quality can affect extraction.
Evaluate the output with a representative sample of your own CVs.

Use parsing in CVReaderPro or connect it to your ATS

Choose the workflow that fits your team and level of automation.

In the CVReaderPro interface

Upload a CV, review structured data and continue your recruiting workflow.

Try it free with 200 credits

In an ATS or application

Use the API to connect extraction to your ATS, HR software or infrastructure.

Compare API options

How much does rule-based CV parsing cost?

The rule-based mode uses fewer credits than AI modes and lets you test a meaningful volume during the trial.

Rule-based parsing: 1 credit per processed CV.
AI parsing 56: 3 credits per CV. AI parsing 96: 5 credits per CV.
The trial includes 200 credits for 14 days, with no credit card required.
If you use all 200 credits only for rule-based parsing, you can process up to 200 CVs.
Test several CVs representative of the documents your organization receives.

CV parsing APIConnect structured extraction to your ATS or business application.CV and job matchingCompare profile data with the requirements of a role.CVs in your company formatPrepare candidate profiles in a template that fits your visual identity.

Frequently asked questions about rule-based CV parsing

What is the difference between rule-based and AI parsing?

Rule-based parsing uses deterministic extraction and normalization, costs 1 credit per CV and provides 56 available fields. AI modes use a model to interpret more context: AI 56 costs 3 credits and AI 96 costs 5 credits per CV.

Can manual corrections and AI results improve the engine?

Validated manual corrections and AI parsing results can help identify cases to improve. Engine changes are reviewed and decided separately; they are not applied automatically after every correction.

Are all 56 fields always populated?

No. 56 is the number of fields available in this mode. Populated fields depend on the information present and usable in the CV.

Which CV formats can I test?

Public demo flows accept files including PDF, DOC, DOCX and TXT. Test your own documents during the trial to check fit with your workflow.

Can I connect parsing to my ATS?

Yes. The API lets you use structured data in an ATS or another application. Visit the API page and documentation for integration details.

How many CVs can I process during the trial?

The trial includes 200 credits for 14 days. Since rule-based parsing costs 1 credit per CV, you can process up to 200 CVs if you use all credits for this mode.

Does rule-based parsing use AI?

No. This mode uses extraction and normalization rules. An AI model is not called for each document processed in rule-based mode.

Try rule-based parsing on your CVs

Upload different document types, review the structured data and choose the mode that fits your volume, CVs and workflow.