Author: Paul Derstine
by Vic Diloreto, Director, Software Products, Elder Research
The Risk Assessment Data Repository (RADR) is a powerful risk analytics platform used to enhance productivity in the investigation of fraud, waste and abuse. This server-based, data analytics product fuses data from multiple sources, with sophisticated predictive and machine learning risk modeling, and an intuitive visual interface. RADR enables proactive identification of risk—namely fraud, waste, and abuse behaviors—and simplifies the investigative process. RADR provides visualizations for risk propensity and their related data so that managers, auditors, investigators, and analysts can easily access data on high-risk items and focus on the highest ROI cases.
This customizable tool combines Elder Research’s world-renowned data science expertise, data aggregation, and multi-layered visualization to enhance productivity in fraud, risk, and threat investigations. Predictive and machine learning models score attributes based on specific risk indicators. RADR isn’t just a tool to build with — it’s a turnkey solution to realize results immediately.
Key Benefits of RADR include:
- Improves analyst/investigator efficiency by consolidating and analyzing data from multiple sources in one tool
- Sophisticated predictive and machine learning models reduce false positives and increase the number and quality of actionable cases
- Visualization and geospatial mapping can visually identify emerging patterns
- Link analysis can aid in identification of organized criminal networks by highlighting hidden connections
- Data export and print capabilities simplify reporting
- Drill down by area, case or entity, map and list views, cost trending, and custom filtering to simplify investigations
- Configuration of the tool is profile based and for archival and version control
Focusing on higher-risk items increases the return on investment of labor hours by investigators, auditors, and analysts. For every labor hour invested, more issues that need to be addressed are likely to be discovered and will therefore accelerate efforts to reduce waste and abuse at a lower cost.
Target applications for RADR include:
- Contract Fraud
- Financial Accounting Fraud
- Healthcare Insurance (Claimant/Provider)
- Dental Insurance
- Life Insurance
- Property & Casualty Insurance
- Automobile Insurance
- System Behavior Analysis
- Worker Compensation & Disability Insurance
Underlying models conduct community detection and network analysis to statistically predict and uncover fraud networks for further investigation and RADR Link Analysis Explorer allows rapid exploration of results.
RADR supports several deployment scenarios:
RADR and Pre-developed Elder Research Models
RADR and Contracts/US Spending Data (RADR-5001)
Elder Research has developed a Contract assessment model that we have applied to the publicly available US Spend data. Based on years of contract anomaly experience, Elder Research has applied this experience to the contract information for all federal agencies supported by USASpending.gov. Elder Research’s model employs several techniques to rate contracts for potential anomalies and incorporates them all into an easy to understand score. The higher the score, the more potentially anomalous the contract could be.
RADR with Model Development Needs (RADR-1000)
RADR-1000 is used in cases where a client does not have an existing model. In this case, we work with the client to develop a model and use RADR to provide the insight and visualization. In one example, Elder Research worked with the U.S. Postal Service OIG to develop a healthcare fraud model. RADR was introduced to strategically visualize the model results, leading to the initiation of 113 investigations and aiding over $9.5 million in recoveries, restitutions, and cost avoidance. The RADR tool amplified the productivity of USPS investigators by reducing the number of hours spent on a case by 30% and increasing the dollars returned per case by 35%.
RADR with Pre-Developed Client Models (RADR-3000)
RADR-3000 is used in cases where a client has data and a model but needs a server-based visualization tool designed to provide investigative insights. In this scenario, RADR is deployed by the client following our on-boarding procedures. As noted in the system requirement section, RADR supports several deployment modes – premise, cloud, Windows, Linux, and more. As a fully supported product, the client has access to administrator and us¬er tutorials and documentation.
This workload prioritization tool enables financial managers, auditors, investigators, and analysts to easily access data on high-risk transactions and focus on the cases with the highest projected return on investment.