Skip to main content

How SeekOut Diversity Classifiers Work

SeekOut Diversity features are designed to help recruiting teams take action to improve organizational diversity by finding and including h…

T
Written by Tirlochan Arora
Updated today

Using SeekOut Diversity to Build Inclusive Talent Pipelines

SeekOut Diversity empowers recruiters to improve organizational diversity by identifying and engaging qualified candidates from underrepresented groups. With advanced filters and machine learning classifiers, SeekOut helps teams build more inclusive pipelines efficiently and accurately.


Diversity Filters

SeekOut offers dedicated filters to help recruiters find candidates who are likely members of the following demographic groups:

  • Women

  • Black

  • Hispanic

  • Asian

  • Veterans

These filters are designed to surface candidates who are statistically likely to belong to these groups based on profile attributes.


How SeekOut Classifies Diversity

SeekOut uses machine learning classifiers trained on tagged data to estimate a candidate’s diversity status. These classifiers analyze a wide range of profile attributes, including:

  • Names

  • Educational background

  • Group memberships

  • Keywords and phrases used on profiles

  • Other publicly available signals

Each model is rigorously tested and retrained regularly to ensure high accuracy. We focus on two key metrics:

  • Precision: How accurately we identify a candidate’s diversity status

  • Recall: How many qualified candidates we’re able to find

Currently, our classifiers achieve over 90% precision and recall, meaning they are both highly accurate and comprehensive.


Why SeekOut Is More Effective Than Manual Methods

Traditional approaches to sourcing diverse candidates—such as searching by common first names—are limited and often introduce bias. For example, searching for the top 50 female names in the U.S. tends to return mostly Caucasian candidates due to name distribution.

SeekOut’s approach is far more inclusive. Our women classifier, for instance, uses over one million names along with other profile attributes, enabling recruiters to find a broader and more diverse pool of candidates.


Reducing Unconscious Bias

To help teams focus on skills and experience, SeekOut offers a Bias Reduction Mode. This feature redacts candidate photos, names, and other identifying details, allowing recruiters to evaluate candidates more objectively.


Customization Options

SeekOut’s diversity filters can be tailored to meet the unique needs of enterprise customers. Customization options include:

  • Adding new filters for hard-to-find groups (e.g., Native American, candidates with disabilities)

  • Modifying existing filters to align with internal policies

  • Limiting or removing filters based on organizational requirements


If you have questions about configuring diversity filters or want to explore customization options, please contact your SeekOut representative or [email protected].

Did this answer your question?