The Accelerator for Fraud, Waste and Abuse shows you which submissions for your program benefits are potentially fraudulent. In order to see possible fraud in your own program, you need to analyse prior periods of data to establish the KPI thresholds and other indicators for potential fraud. 
Avaap is here to make you successful.

### Answer Key Business Questions
- What’s the financial impact of potentially fraudulent applications?
- How many applications did we receive over the last three months?
- What was the total amount disbursed to these applicants?
- What geography/demographics trends do we see in potentially fraudulent applications?
- What percentage of applicants used a bank account, email address or phone number previously used by at least one or more other applicants?

### Monitor and Improve KPIs
- Application volume by state
- Likelihood of fraudulent application
- Duplicate bank account submissions
- Duplicate IP address submissions
- Duplicate email submissions
- Duplicate address submissions
- Duplicate phone submissions

### Required Data Attributes
Data Source: Program Summary
- Predictive Score Range (string)
- Historical Count of Fuzzy Email (string)
- Historical Count of IP Address (string)
- Historical Count of Phone (string)
- Application Sub-Type (string)
- Application Submit Day (date)
- Application Type (string)
- Benefit Programme (string)
- Email  (string)
- Fuzzy Email  (string)
- IP Address  (string)
- Last Paid Date (date)
- Person ID (string)
- Phone Number (string)
- Predicted Disposition (string)
- Predicted Score (int)
- Person ID (string)
- State (string)
- Total Amount Paid (string)

Data Source: Daily Summary
- Application Submit Day (date)
- Benefits Programme (1/2) (string)
- Daily Applicant count by Programme (int)
- Daily Duplicate Address Count (int)
- Daily Duplicate Bank Count (int)
- Daily Duplicate Email Count (int)
- Daily Duplicate IP Count (int)
- Daily Duplicate Phone Count (int)


Fraud, Waste, and Abuse

Monitor and predict fraudulent submissions for claims benefits