Software company implements process to safeguard personal data
Home inspection software company engages Baker Tilly to institute a process to identify and redact customer data containing personal identifiable information. Read this case study to learn more.
Client background
The client provides leading edge home inspection software and internet tools that inspectors use to deliver comprehensive and easily accessible property information to home and commercial buyers, sellers and real estate professionals.
The business challenge
The client launched a new online business unit focused on the insurance industry, collecting property inspection data as part of a fraud prevention and risk stratification strategy. However, the data contained both sensitive and non-sensitive PII (personally identifiable information), including names, addresses, phone numbers, social security numbers, driver’s license numbers, credit card details and other sensitive personal data. This posed a significant challenge as they needed to ensure this information was protected before it could be leveraged for customer behavior research or shared with offshore development partners.
Strategy and solution
Baker Tilly worked closely with the client to develop and implement a robust data redaction process that ensured the protection of PII in all documents and images containing sensitive information. The strategy was designed to address both textual and visual data:
- Textual redaction: A sophisticated automated process was created to identify and redact PII in text documents.
- Image redaction: Images containing personal characteristics, such as faces, face geometry and fingerprints, were also identified and redacted. Artificial intelligence (AI) technology was used to automatically detect and redact these visual data points.
- Human review process: Low-confidence redaction decisions, where automation was uncertain about the accuracy, triggered a manual review by human experts. This hybrid approach helped maintain the highest standards of data protection and ensured precision in identifying sensitive data.
- Continuous improvement: Data from human reviews was fed back into the system, allowing the automated redaction tool to evolve and improve over time, further streamlining the process and increasing the overall accuracy.