Mr. Proper is a multi-subnet AWS VPC architecture that anonymizes sensitive corporate data before ingestion into AlexandrIA, TYPSA's enterprise AI platform. Data flows through five specialized subnets: Ingestión normalizes documents via Lambda, Clasificación uses Google ADK to categorize content, Anonimización applies Presidio and spaCy for PII removal, Integración y Trazabilidad logs transformations to S3 (KMS) and RDS PostgreSQL, and Deanonimización reverses masking for authorized LLM access. The design enforces least-privilege access through VPC isolation, KMS encryption, CloudWatch monitoring, and CI/CD security scanning via GitHub Actions and CodeBuild, with an on-premise air-gapped semantic firewall for additional control. Fork this diagram to customize subnet CIDR ranges, add additional classification scenarios, or adapt the anonymization pipeline for your own regulated data workflows.