How the Commonwealth of Kentucky Supercharged Employee Productivity Using QDox and AWS
Executive Summary
The Commonwealth of Kentucky’s Department of Education and Labor Cabinet receives thousands of citizen unemployment benefits and insurance application claims every month. Historically, processing these documents involved significant manual effort, leading to inefficiencies and challenges in tracking insights. By leveraging Quantiphi’s GenAI-powered intelligent document processing solution, QDox, the agency automated document classification, data extraction, and PII redaction. This transformation successfully processed 10,000 forms per month, saved 1,600 man-hours monthly, and achieved a 97% automated review rate—greatly improving operational efficiency, data security, and fiscal visibility.
About the Client
The Commonwealth of Kentucky’s Department of Education and Labor Cabinet is a US State Government Agency responsible for protecting the rights and welfare of wage-earning employees. The agency handles substantial volumes of citizen requests, including unemployment benefits and insurance application claims, ensuring that essential public services are delivered securely and efficiently.
Problem Statement
The agency’s manual process for managing high volumes of citizen unemployment benefit and insurance application claims required significant administrative time and effort. As the volume of documents grew, the agency sought an AI-powered solution to automate document classification, streamline data extraction, and redact personally identifiable information (PII) to ensure data security and improve processing times.
Challenges
The agency’s existing document operations presented several critical hurdles:
- High Volume of Manual Processing: Staff faced a heavy manual workload reviewing, extracting data from, and processing thousands of incoming citizen benefit and insurance claims forms.
- Data Security and Privacy Risks: The presence of sensitive Personally Identifiable Information (PII) in applications required strict, time-consuming oversight to ensure compliance and data privacy.
- Lack of Actionable Insights: The agency struggled to easily gain insights into document processing volumes, extraction rates based on specific time periods, and peak processing times.
- Communication Bottlenecks: There was no system in place to automatically notify relevant departments when the required fields from submitted documents had been updated, causing delays in cross-departmental workflows.
Solution
Quantiphi implemented QDox to automate and secure the agency’s document processing lifecycle. To deliver a seamless and efficient experience, we enabled:
- Intelligent Document Processing (IDP): Deployed QDox, powered by Amazon Textract, to automatically classify incoming forms and extract essential information from diverse insurance and unemployment claim documents.
- Automated PII Redaction: Integrated automated redaction capabilities to securely identify and mask Personally Identifiable Information (PII), ensuring strict data security standards are maintained.
- Advanced Analytics & BI: Developed a comprehensive BI dashboard to provide leadership with actionable insights into document processing trends, benefits allowances rolled out, beneficiaries served, and claim payout statuses.
Results and Business Impact Created
- High-Volume Scalability: Successfully processed 10,000 citizen benefit forms per month through automated document classification and data extraction.
- Massive Efficiency Gains: Saved 1,600 man-hours monthly, significantly reducing manual workloads and freeing staff to focus on complex citizen services.
- Touchless Automation: Achieved a 97% automated forms review rate, requiring zero human intervention for the vast majority of submissions.
- Enhanced Data Security: Ensured strict compliance and protection of citizen data by accurately and automatically redacting sensitive PII from incoming forms.
- Actionable Visibility: Empowered leadership with a BI dashboard offering real-time insights into citizen unemployment benefits, total allowances, monthly beneficiaries, and claims payout versus denied claims.
AWS Services Used
Amazon Textract
Amazon SageMaker
AWS Lambda
Amazon S3
Amazon RDS
Amazon Route 53
Amazon DynamoDB
Amazon SQS