Background checks are a vital component of the federal government, especially with so many new appointments during the administrative transition. All federal employees (and many commercially employed contractors) are required to submit personal information in the form of an application and undergo some level of background investigation. In fact, a new federal agency within the Office of Personnel Management called the National Background Investigation Bureau (NBIB) was recently created to accommodate the increased volume and ensure information security for the millions of applications processed each year. Of course, the applications represent only a small percentage of the documents reviewed during the vetting process.
One of NBIB’s key new initiatives is the establishment of a Federal Investigative Records Enterprise (FIRE) office to automate and manage government-wide information sharing and standardization, a difficult task given the unstructured nature of documents. The NBIB would greatly benefit from leveraging document analytics and machine learning technology like Ephesoft Insight to bring the background investigation process into the 21st century.
The primary goals for NBIB federal background investigations are determining the (1) eligibility for logical and physical access, (2) employment suitability and fitness, and (3) eligibility for access to classified information or to hold a sensitive position. While these criteria may appear straightforward, the path to achieving consistency within the standards of investigation and federal agency reciprocity has been fraught with challenges.
Consider volume from a big data perspective: an enormous amount of personal information is required to validate an individual for employment and security clearance. While a significant percentage of information is provided voluntarily by the applicant (via application forms SF85, SF86, etc.), most of these details require validation from external and third-party data sources like credit agencies, educational institutions, Customs and Border Patrol logs, federal, state and local criminal record databases, and so on. The list is exhaustive, and data formats are highly disparate. For example, state and local ordinances for records management and information storage vary greatly, and querying that information for review is an unstandardized process. Similarly, verification of air travel is relatively simple while sea travel verification introduces difficulties due to the nature of port records. This diversity presents a substantial validation challenge to the Bureau.
The sheer quantity of documentation that the NBIB manually collects for a single applicant is staggering. Ephesoft Insight custom crawlers would enable record queries from these external data sources to be automatically incorporated into a background investigation file. Rather than active transmission of electronic or paper records from agency to agency, Ephesoft Insight could pull information from third-party repositories and databases to compare against application data fields. This relational analysis automation for information verification will supply investigators with the context required to quickly make clearance grant or deny decisions.
Now consider the often overlooked value of the data that lives within the pages of all applications and supporting documentation submitted to the US Government over the past 25 years. While external databases and data sources may be referenced for information verification, the existing pool of historic applicant data is an untapped resource. Ephesoft Insight could aid the NBIB in the investigation process and handily identify potential red flags by holistically considering all historic applications and normalizing the unstructured content.
For example, there is inherent value in quickly identifying whether an applicant has been previously denied security clearance. With investigations costing the government from $200 to $5,000 (depending on the investigation tier) the ability to promptly deny an applicant that submits multiple applications (whether purposefully or unintentionally) before starting the background check process would save significant resources. Ephesoft Insight could identify duplicate applications by unique data points like social security number, name or date of birth to flag these incidents through a dashboard visualization.
Finally, consider the application-required data points of foreign contacts and previous residences. When reviewed individually, applicants that list foreign contacts that are on a U.S. or global security watch list will likely be denied clearance. However, NBIB investigators do not always have the benefit of analysis to compare previously denied applicant residence history with new and ongoing background investigation information. Consider that an applicant may not voluntarily list all known foreign contacts. However, he or she may have lived at the same address over the same period as a previous applicant who was denied clearance because of a relationship with a flagged foreign contact. Relationship analysis in this context will arm investigators with more complete information with which to make informed decisions.
Looking at the background investigation process as a data-driven operation instead of an individual approval process will aid the OPM and the United States Government in conserving time and resources, saving money, and protecting our country’s valuable intelligence and information. With key challenges of applicant vetting through NBIB program management, including a disparity of applicant information, non-standardized adjudication processes across agencies, and a prolonged timeline for investigation, Ephesoft Insight would aid in the standardization and automation of information gathering, application information validation through relational data analysis, and holistic applicant data review.
We’re ready to help.