Monitoring grantee performance is crucial for the success of every grant program. Grantee risk assessments and performance data are often collected and stored within bureau or office-level silos inside an agency. This can lead to redundancy of effort and a lack of transparency, ultimately increasing the potential for negative risks.
REI utilized machine learning to develop a Risk Management Framework for assessing grantee risk. A weighted formula was applied, taking into account numerous factors that influence risks, in order to establish the framework. A risk score can be determined for each grant or grantee. By evaluating single audit data alongside historical outcomes, such as drawdown restrictions, the relative risk associated with award decisions can be determined during the grant evaluation process. This risk prediction model is trained on over a dozen audit data elements, including reportable conditions, material non-compliance, and material weakness, as well as awards data elements. REI employed machine learning to create a Risk Management Framework to support grantee risk assessment.
Establishing a shared understanding of risk factors and employing a consistent scoring mechanism promotes consistency among bureaus and offices within an agency, eliminating redundancy and re-work. Applying AI to grants management systems allows for the automatic flagging of high-risk applicants during both the grant application phase and the post-award phase. This can be utilized to prioritize tasks for grants officers and support informed decision-making.
- Multi-Factor Weighted Analysis
- Machine Learning
- Risk Management Framework