CV Parsing & Screening Testing
Bias auditing and accuracy testing for CV parsing tools and automated resume screening systems.
The Challenge
CV parsing and screening tools sit at the heart of modern recruitment. They process thousands of applications, deciding which candidates move forward and which are filtered out. But these systems often encode the very biases they're supposed to eliminate.
Amazon's infamous recruiting AI—which taught itself to penalize resumes containing the word "women's"—is just one example. Similar biases exist in commercial tools today, often hidden in complex feature weights that even vendors don't fully understand.
We test CV screening systems to find these hidden biases and help you build fairer processes.
What Can Go Wrong
Common bias patterns we test for in CV parsing and screening systems
Gender Bias
CV parsers may penalize terms associated with women (e.g., "women's basketball") or favor traditionally male-coded language.
Name Discrimination
Studies show CVs with ethnic-sounding names receive systematically lower scores from AI screening tools.
Age Inference
Graduation dates, work history length, and technology skills can be used to infer and discriminate based on age.
Education Bias
AI may over-weight prestigious universities or penalize non-traditional educational paths.
Format Sensitivity
CVs in non-standard formats may be incorrectly parsed, causing qualified candidates to be rejected.
Gap Penalization
Employment gaps (often due to caregiving) may be unfairly penalized by screening algorithms.
Our Testing Methodology
A rigorous approach to uncovering bias in CV screening systems
Parsing Accuracy
Test how accurately the system extracts information from diverse CV formats.
Bias Testing
Submit matched CV pairs varying only in protected characteristics.
Proxy Detection
Identify features that serve as proxies for protected characteristics.
Statistical Audit
Analyze historical screening data for discriminatory patterns.
What You Get
Bias Report
Quantified evidence of any discriminatory patterns with statistical analysis.
Accuracy Assessment
Evaluation of parsing accuracy across different CV formats and structures.
Mitigation Strategies
Practical recommendations for reducing bias while maintaining efficiency.
Is Your CV Screening Fair?
Find out what biases might be hiding in your resume screening process.
Request Assessment