Leveraging the Power of Machine Learning for Better Background Checks

While other consumer reporting agencies are forced to process every record manually, our machine learning tools categorize charge data clearly and consistently. This makes the hiring process faster and safer, allowing Checkr to compliantly optimize candidate reports, as well as automate some hiring processes based on your preliminary assessment rules. Processing over 1.5 million background checks per month, Checkr's platform is constantly learning, yielding higher-quality results, and giving our customers greater clarity and control.

  • Clear and accurate charge reporting
  • More consistent hiring decisions
  • No lag in processing time, no matter the volume

What Machine Learning Does for Background Checks

Clear, Consistent Charge Data

Our Charge Classifier uses natural language processing to quickly categorize criminal charges from jurisdictions across the U.S., drastically reducing the need for manual review.

Faster Turn-Around Times

With standardized data from our Charge Classifier, Checkr's platform can deliver more accurate candidate records in less time.

Multiple Tiers of Quality Assurance

Constant audits and evaluation improve our machine learning models to get the efficiency of automation along with the confidence of targeted human review when necessary.

Scalability to Meet Demand

While other background check providers' output can lag when volume increases, our technology can process 20 million reports a year with no impact on turn-around time.

ebook

The Beginner's Guide to Background Checks

Blog

How Checkr is Leveraging Humans to Build Trust in Technology

Meet Checkr's Charge Classifier

Our language processing algorithm has a growing body of thousands of words in its vocabulary, and it gets smarter with every background check we process. This is important because records data is written by humans, so it can be inconsistent. For example, our Charge Classifier has found over 200 unique ways to report that a charge has been “dismissed.” When the Charge Classifier sees any of those 200 versions of a dismissed charge on a background check, it automatically standardizes the data for faster, more error-free reporting. Easy-to-read reports built through consistent classification make the adjudication process fairer. And, this standardized data paves the way for our hiring tools to help you adjudicate faster and more compliantly.

Combining the Strength of Technology With the Expertise of Data Engineers

Machine learning helps Checkr standardize language for greater consistency and greater speed, while our Quality Assurance team constantly maintains and reviews the results for efficiency and accuracy. Checkr's machine learning creates powerful efficiencies, but human review also plays an important role. When our machine charge classifier thresholds are not met, a criminal record quality specialist will manually review and determine the outcome. In addition, certain types of records are always manually reviewed for added safety. This, coupled with routine audits, creates a continuous quality feedback loop to keep our models compliant and accurate.

Learn More About Checkr's Technology

Our tools update along with changing regulations and surface only legally reportable records to help keep your hiring compliant.

The right data can help you keep your workplace and customers safer. Keep tabs on new records connected to your candidates and employees with our Real-Time Feed.

Ready to Give Us a Try?

We've built the most advanced background check toolset in the industry to make hiring faster, easier, more accurate, and lower risk.