Candidate screening

Automate Candidate Screening with Structured Resume Data

Screen every applicant against your exact criteria in seconds, not days

Try screening automationView documentation
The problem

A single job posting generates 150-300 applications, and your recruiters spend 6-8 seconds per resume during initial screening. At that pace, qualified candidates are missed because reviewer fatigue sets in after 50 resumes, and screening criteria drift throughout the day. Worse, different recruiters apply different standards to the same role, so your shortlist depends more on who reviewed the resume than on the candidate qualifications. When you are hiring for 20 roles simultaneously, inconsistent manual screening is the biggest source of lost talent.

The solution

Resume Parser extracts skills, job titles, employment durations, education levels, and certifications into structured JSON. You define screening rules against this structured data: knockout criteria that immediately disqualify (no required certification, insufficient experience), must-have skills that candidates need to proceed, and nice-to-have skills that boost their ranking score. Your screening logic runs against the parsed output, scoring and ranking every applicant identically. Recruiters receive a shortlist ranked by fit, with the parsed data justifying each candidate score.

How it works

1

Define screening criteria per role

For each open position, specify knockout criteria (disqualifiers like missing a required license), must-have skills with minimum experience thresholds, and nice-to-have skills with weighted scoring. Store these as structured rules that your screening engine evaluates against parsed resume data.

2

Parse every incoming application

As resumes arrive, send each to the Resume Parser API. The synchronous response returns structured JSON in under 8 seconds, including extracted skills (both technical and soft), work experience entries with durations, education with degree levels, certifications, and languages.

3

Apply knockout and ranking logic

Your screening engine first checks knockout criteria: does the candidate have the required certification? Do they meet minimum years of experience? Candidates who pass knockouts are scored on must-have skills (weighted heavily) and nice-to-have skills (weighted lightly). The output is a ranked list with a numeric score and a breakdown showing which criteria each candidate met or missed.

4

Recruiter reviews the ranked shortlist

The recruiter opens a ranked candidate list sorted by screening score. Each entry shows the overall score, which knockout criteria were met, which must-have skills matched, and which nice-to-have skills boosted their ranking. The recruiter focuses evaluation time on the top-ranked candidates rather than re-reading every resume.

Key benefits

Screen 300 applicants with the same rigor as the first 10

Human screeners lose consistency after reviewing 40-50 resumes. Automated screening applies your exact criteria to every single applicant, whether they are number 1 or number 300. This means qualified candidates buried at the bottom of the pile get the same fair evaluation as those at the top.

Separate knockout criteria from preference scoring

Not all requirements are equal. A missing nursing license is a hard disqualifier, while Python experience might be a strong preference. Structured screening lets you define knockout criteria that eliminate unqualified candidates immediately, then score remaining candidates on a weighted mix of must-have and nice-to-have skills.

Reduce unconscious bias with criteria-based evaluation

When screening decisions are based on parsed skills, experience durations, and certifications rather than resume formatting, school prestige, or name recognition, you reduce the influence of unconscious bias. Every candidate is evaluated against the same objective, job-relevant criteria that you defined before applications arrived.

Surface qualified candidates you would have missed

Automated screening catches candidates whose resumes are formatted unconventionally or who describe skills using different terminology. Because the parser normalizes extracted data, a candidate who lists "React.js" and one who lists "ReactJS" both match your "React" requirement. Manual screeners often miss these matches during rapid review.

Frequently asked questions

Knockout criteria are binary disqualifiers tied to hard requirements: the candidate either has a required certification or they do not, they either meet minimum experience years or they do not. Must-have skills are strongly weighted in scoring but do not automatically disqualify. Nice-to-have skills add bonus points. A practical setup for a senior DevOps role might be: knockout if less than 5 years in IT, must-have AWS and Kubernetes, nice-to-have Terraform and Ansible. You store these rules as structured data and evaluate them against the parsed JSON fields.
You build the ranking algorithm using the structured JSON output. A simple approach: assign weights to each criterion (e.g., must-have skill match = 10 points each, nice-to-have = 3 points, extra experience year = 2 points). Sum the weighted scores for each candidate after knockout filtering. More sophisticated approaches calculate skill overlap percentages, penalize experience gaps, or boost candidates with rare skill combinations. The parser gives you the structured fields; your logic defines the scoring.
Automated screening is legally permissible when criteria are job-relevant and applied consistently. The key is that your knockout and scoring criteria must reflect genuine job requirements, not proxies for protected characteristics. Document your criteria, apply them uniformly, and periodically audit outcomes for adverse impact. Resume Parser extracts objective data like skills, certifications, and experience durations, which are appropriate bases for screening decisions.
Build a skills synonym table that maps variations to canonical names. The parser extracts skills as written on the resume, so "JavaScript", "JS", "ECMAScript", and "Javascript" might all appear. Your screening logic should normalize these before matching against criteria. Start with the 50-100 most common skills for your industry and expand as you encounter new variations. This normalization step is critical for accurate screening at scale.
The parser extracts work experience entries with job titles, companies, and date ranges. You can calculate total experience and infer skill-specific experience by associating skills with the roles where they were likely used. For example, if a candidate lists "Kubernetes" as a skill and their last two roles (spanning 4 years) are DevOps Engineer positions, you can reasonably attribute 4 years of Kubernetes experience. Exact skill-to-timeframe mapping requires your business logic on top of the parsed data.

Related pages

Use cases

Resume Parsing That Powers End-to-End HR Automation

Use parsed resume data to trigger onboarding workflows, populate HRIS records, and enforce compliance checks. REST API with structured JSON output for workflow automation platforms.

Features

Skill Extraction API

Extract skills from resumes with 99.2% accuracy. Skills are identified in context from work history and mapped to roles, not just keyword-matched.

Use cases

ATS Resume Parsing That Populates Candidate Profiles Automatically

Parse resumes into structured JSON and map fields directly to your ATS. Works with Greenhouse, Lever, Workday, BambooHR, and iCIMS via REST API with 99.2% field accuracy.

Solutions

Resume Parser for Staffing Agencies

Parse thousands of resumes daily with 99.2% accuracy. Reduce time-to-fill with automated candidate data extraction, ATS-ready JSON output, and VMS-compatible formatting. Built for high-volume staffing operations.

Build your screening criteria against real parsed data

Parse 10 resumes from a recent job posting and prototype your knockout and ranking logic against the structured output. See how automated screening compares to your manual shortlist.

Try screening automation