Skyflow for PostgreSQL Demo

A simple demonstration of using Skyflow's tokenization with PostgreSQL using BEFORE INSERT/UPDATE triggers.

Features

• Automatic tokenization of PII data (name, email, phone, address) on insert/update
• Detokenization support using SKYFLOW_DETOKENIZE function
• Demonstrates both single row and batch operations
• Uses PostgreSQL's trigger system for seamless integration
• Direct HTTP requests to Skyflow API using PostgreSQL's HTTP extension

Requirements

• Homebrew (for installation)
• Skyflow credentials:
  • Account ID
  • Vault ID
  • Auth Token/API Key

Vault Creation

Before proceeding with the setup, you'll need to create a vault in Skyflow:

1. Log into your Skyflow account
2. Create a new vault using the included vault schema file (vaultSchema.json)
   • This schema defines a table for storing PII data with appropriate tokenization policies
   • The schema includes a pii table with fields for storing sensitive information
   • Tokenization is configured using DETERMINISTIC_FPT policy, for deterministic + format-preserving tokens.
3. Once created, note down the Vault ID as it will be needed during setup

Setup

Use the provided setup script to manage the PostgreSQL environment:

# Create environment
./setup.sh create

# Destroy environment
./setup.sh destroy

# Recreate environment
./setup.sh recreate

# Show usage
./setup.sh

The create/recreate commands will:

• Install and configure PostgreSQL 17
• Install required extensions
• Prompt for your Skyflow credentials
• Set up the demo database and tables

Example Usage

-- Insert a customer (PII fields will be automatically tokenized)
INSERT INTO customers (
    name, 
    email, 
    phone, 
    address, 
    lifetime_purchase_amount, 
    customer_since
) VALUES (
    'John Smith',
    'john@example.com',
    '555-222-5555',
    '123 Fake Street NY NY 10019',
    5000,
    '2020-01-01'
);

-- Query the data to see tokenized values
SELECT * FROM customers;

-- Query with detokenization
SELECT SKYFLOW_DETOKENIZE(name), SKYFLOW_DETOKENIZE(email), customer_since 
FROM customers;