A criminal case involving export control
Total Time to Generate, Report, and Act on Results = 3 Days
VortexT was employed after a five-year investigation by a Federal intelligence agency to analyze case data residing in 35 hard drives containing 2.7 million documents. Time was of the essence as the prosecution was in process.
- Loaded data on a desktop computer overnight.
- Data analysis and visual representation of the linkages in the case data: 8 hours.
- Reviewed findings with case agents and verified main points of the case.
- Discovered previously unknown facts: new suspects, an active shell company, the target’s organizational details, etc.
A long-term care claims analysis at a Top 5 U.S. insurer
Total Time to Produce 832% Return on Investment (ROI) = 16 hours
The insurance company processed a total of 26,000 claims annually generating $900M of claims payments. Each paid fraudulent or questionable claim cost the insurance company approximately $200K. The client employed hundreds of trained professionals, structured data tools, predictive models, and third-party service bureaus to identify fraudulent or questionable claims. Claim determination involves human analysis of hundreds, if not thousands, of pages of unstructured text (e.g., telephone transcripts, emails, home visits, third-party assessments, notes from multiple systems, etc.) VortexT ingested a data set of 20,000 claims on a laptop in minutes.
- Revealed additional instances of fraudulent or questionable claims paid and in process over those previously known by the company.
- Discovered new “red flag” for profile of fraudulent claims.
Analyses of radiology reports
Time to Produce Results = 2 days
VortexT analyzed radiology reports of unstructured text for 300 patients with abdominal aortic aneurysms (AAA) to determine why some patients had post-surgical complications with no clear indication as to the cause and if there were precursors to the successful or unsuccessful outcome of a surgery. Without evidenced-based information, doctors could only rely on prior experience based on a patient’s pre-surgical condition.
- Revealed comorbidities (i.e., sigmoid diverticulosis and gallstones) through visualization of radiology reports for patients with post-surgical complications.
- Resulted in a changed approach to patients with AAA with comorbidities to improve patient outcomes—a major reimbursement factor under the Affordable Care Act.
VortexT also analyzed large amounts of structured and unstructured data to overcome a significant challenge in the consistency and accuracy of radiologists’ reports of image results. While portions of these reports are sometimes standardized, radiologists document their observations using natural language.
- Results of this evaluation showed normal reports tend to use the same terminology and were very consistent, while abnormal reports use a wide range of terminology and were inconsistent.
- Recommended more standardization in the description of abnormalities identified by the radiologists to aid radiologists in identifying trends in a patient’s health that may improve preventative identification of cancer or other diseases.
Data discovery within a targeted market
9,000 RSS Feeds Automatically Processed Daily Using the Mobile Recommender
VortexT is a data discovery tool used in the time-sensitive IT and software development market—an integration of business intelligence and big data analytics.
- Automated daily download and analysis of over 9,000 user-selected RSS feeds, including news and publications, grants, requests for proposals, etc.
- Using a profile document consisting of 20 papers published by the users, VortexT’s Recommender delivers rank-ordered opportunities for funding, grants, and contracts with speed and accuracy in order to make better informed business decisions.