Personal identifiable information hiding in your data stack is one of the most common compliance risks in data engineering, and one of the hardest to find manually. Most teams don't have a complete picture of which columns contain names, emails, phone numbers, or other PII across their dbt models and raw source tables.
This demo shows Altimate Code scanning a full Airbnb dbt project running on Snowflake to find PII, then building a GDPR-compliant masking implementation without any manual file editing. Before writing a single line of code, the agent asks about compliance requirements: whether names should be hashed or tokenized, whether review text should be scrubbed for embedded PII, and what to do with columns that are high-risk but ambiguous.
The implementation creates a reusable mask_pii_text macro using regex to strip emails, phone numbers, and URLs from free text. Names are masked with SHA-256 hashing, deterministic, so joins and surrogate keys remain consistent. After changes, the agent runs a full dbt build with refresh and pulls a sample output to visually confirm the masking worked. Total cost: $3.
- →How to discover PII across a dbt project and Snowflake schema automatically
- →How to implement GDPR-compliant data masking using dbt macros
- →How to hash identifiers while keeping surrogate keys consistent for joins
- →How to validate masking with a full dbt build and visual data inspection
- Scans all dbt models and raw Snowflake tables for PII column patterns
- Asks GDPR/CCPA compliance questions before acting: hashing vs. tokenizing vs. dropping
- Creates a reusable mask_pii_text macro with regex for emails, phones, and URLs
- Applies SHA-256 hashing to name columns to maintain join consistency
- Runs dbt build with full refresh and validates output visually, $3 total
