[Hook] Database deployment tools usually have a blind spot: they handle schema changes well, but leave you writing shell scripts or ETL pipelines for config files and seed data. pgmi changes that by loading your entire project into PostgreSQL session tables, letting you handle everything—migrations, data loading, and testing—in standard SQL. [What is this video?] This is an introduction to pgmi, a single Go binary that turns PostgreSQL into your deployment engine. We cover how pgmi eliminates the "deployment impedance mismatch" by treating files as queryable data. [Key Features Covered] • Files as Data: Load JSON, XML, and CSV files directly into and query them with SQL. • Transactional Testing: Run tests inside your deployment transaction using PostgreSQL savepoints. If a test fails, the entire deployment rolls back. • Environment Logic: Use parameters to load specific data for dev, staging, or production without external scripting. • Single Artifact: Replace scattered SQL, shell scripts, and loaders with one project directory and one . [Links] • GitHub: [Timestamps] 0:00 - The Problem: SQL vs. Data Files 1:30 - How pgmi works (Session Tables) 3:00 - Loading JSON & Config in 5:00 - Transactional Testing with Savepoints 7:00 - CI/CD Integration #PostgreSQL #DevOps #DatabaseMigration #Golang #SQL #pgmi