Backend Models

Define database models by specifying data types (string, integer, float), constraints (NOT NULL, UNIQUE, CHECK), and relationships in your schema file. Use ORMs like Prisma, TypeORM, or Drizzle to enforce these rules and ensure data integrity at the database level.

Establish relationships by defining foreign keys and relationship mappings in your ORM schema. Specify one-to-many, many-to-one, or many-to-many relationships, then configure cascade behaviors to control how deletions and updates propagate across related records.

Implement validation rules directly in your schema definition using ORM constraints, and add createdAt and updatedAt timestamp fields automatically. This ensures consistent data validation and tracks record lifecycle without application-level logic.

Each ORM has different schema syntax and features, but all support core concepts: data types, constraints, relationships, and indexes. Choose based on your project's needs; the principles of model design remain consistent across them.

Indexes improve query performance and enforce uniqueness; constraints prevent invalid data from entering the database. Together, they protect data integrity, reduce errors, and make schema evolution safer and more predictable.

Add a deletedAt timestamp field to your model and filter queries to exclude soft-deleted records by default. This preserves data history and audit trails while logically removing records from active use.