Leveraging RPA to Mask 500,000 National ID Numbers with 98% Accuracy for Pramerica Life Insurance
Success Story

About the Client

Pramerica Insurance Logo

Pramerica Life Insurance is a leading life insurance company headquartered in Gurgaon, India. Serving over 20 million customers nationwide, the organization operates through a network of 140+ branches and is supported by a workforce of more than 3,000 employees across India.

Country:

India

Industry:

Insurance

Download PDF

Business Situation and Requirements

Under regulatory guidelines, life insurance companies in India are required to store national ID numbers in a masked format. Meeting this requirement involves updating legacy systems, securely handling sensitive customer data, integrating with existing tools, and maintaining smooth operations across policies and claims. At the same time, organizations must ensure full compliance while safeguarding customer information.

To address these challenges, Pramerica Life Insurance aimed to develop a solution that could automatically mask national ID numbers already present in customer documents. The primary goal was to strengthen data privacy and security, protecting sensitive personal information while still allowing the documents to be used for legitimate business and verification purposes. This solution not only ensures compliance with regulatory mandates but also reduces the risk of unauthorized access to confidential customer data.

Their key requirements include:

  • Automated Data Masking: Implement a solution to automatically mask national ID numbers across large volumes of customer documents.

  • High Volume Processing: Efficiently process approximately 500,000 documents without manual intervention.

  • Integration with IBM FileNet: Seamlessly access and retrieve documents stored on the FileNet enterprise content management system.

  • Secure Replacement of Original Files: Replace original documents with masked, compliant versions while maintaining data integrity.

  • Compliance Adherence: Ensure all masked documents meet the Government of India’s data security mandate.

  • Operational Efficiency: Minimize downtime and optimize performance to support large scale automation without impacting business operations.

Solution

Unthinkable leveraged UiPath’s intelligent automation platform to develop a logic driven Windows based utility for masking national ID numbers. The engagement began with a Proof of Concept (PoC) to validate the solution’s accuracy and feasibility.

Automated Document Retrieval:

UiPath bots were to automatically fetch unmasked image files from Pramerica’s FileNet eDMS, significantly reducing manual effort and ensuring consistent processing across large volumes of documents.

Intelligent ID Detection:

A custom built solution leveraging Google Vision API was developed to accurately identify national ID numbers in scanned images, overcoming challenges posed by inconsistent layouts and manually uploaded documents.

Automated Document Retrieval

Python-Based Processing Scripts:

Custom Python scripts were created to automate the masking process, replace original records with compliant versions, and ensure high efficiency and minimal errors during processing.

Secure On-Premises Deployment:

The entire solution was deployed locally on Pramerica’s desktops, ensuring secure handling of sensitive customer data and maintaining compliance with data protection mandates.

Secure On-Premises Deployment

Impact

Pramerica Life Insurance automated the masking of over 500,000 national ID numbers using RPA, completing the task well ahead of the government’s deadline something impractical manually.

The solution processed each document in about 10 seconds with 98% accuracy, ensuring regulatory compliance, minimizing errors, and safeguarding sensitive customer data. This approach significantly improved operational efficiency and demonstrated the effectiveness of automation for large scale enterprise document management.

Leveraging RPA to Mask 500,000 National ID Numbers

500,000+

documents masked

>98 %

accuracy

5 million+

jobs executed by bots