This paper introduces a Cognitive Policy Engine (CPE) that utilizes advanced AI technologiesto automatically read and analyze social program policies (e.g., SNAP, Medicaid) and generatestructured eligibility rules. These rules are used for the automated evaluation of applicants andofficer-assisted decision support. By leveraging cloud-based NLP services, the CPE performsentity extraction and dependency parsing to convert policy text into machine-readable JSONformat. This enables the seamless evaluation of applicant eligibility and provides humanreadable,explainable outputs for program officers. A proof-of-concept demonstrates thefeasibility of this system on multiple policies, generating structured rules and deliveringactionable insights that improve efficiency and transparency in decision-making.The proposed framework demonstrates how AI-driven policy-to-rule transformation can beoperationalized responsibly for large-scale public benefit administration with human-in-the-loopgovernance.Keywords: Cognitive Policy Engine, Rules-as-Code, Eligibility Systems, Social Programs,Integrated Eligibility System (IES), AI-driven Policy Automation, Google Cloud NLP, OfficerDecision Support, Policy-to-Rule Transformation
Kamini Sahu (Wed,) studied this question.