Organizations often need to use database copies for development, testing, and analytics, but these datasets may contain sensitive information (PII/PHI). Current de-identification processes are manual, slow, and error-prone via some OCI based tools (ex - Data Safe).
This idea proposes an AI-driven agent deployed on Oracle Kubernetes Engine (OKE) that automates the end-to-end de-identification of non-production database copies. The agent analyzes schemas, identifies sensitive data, and generates a policy-driven masking plan. Execution is handled using Oracle Data Safe and OCI Vault to ensure security and compliance.
The solution reduces data preparation time from days to minutes, ensures consistent and auditable masking, and minimizes dependency on manual processes—enabling faster, secure, and scalable use of non-production data.
