ORM

An object-relational mapping (ORM) is a programming technique that links object‑oriented code with relational databases. An ORM provides a layer that maps classes to tables and objects to rows, allowing developers to work with data using native language constructs instead of writing SQL. It handles persistence by translating object operations into SQL queries, managing connections, transactions, and schema evolution.

Core concepts include mapping metadata, relationships between entities, and data loading strategies such as lazy and eager loading. Common patterns are the identity map, which ensures a unique in‑memory instance per row, and the unit of work, which batches changes for efficient persistence. ORMs often offer migrations, validation, caching, and concurrency controls.

There are different architectural approaches. The active record pattern combines domain objects with persistence methods, while the data mapper pattern keeps domain logic separate from persistence. Many ecosystems provide popular frameworks, such as Hibernate and JPA (Java), SQLAlchemy and Django ORM (Python), Entity Framework (C#), Eloquent (PHP), and TypeORM (TypeScript).

Benefits of ORM include increased developer productivity, a uniform API across data stores, and reduced boilerplate. Drawbacks can include overhead, abstraction leakage, and performance pitfalls such as the N+1 query problem, complex joins, or the temptation to hide expensive queries behind high-level calls. Effective use often involves understanding when to rely on the ORM and when to write bespoke SQL for critical paths.

ORMs are valuable when the domain model maps well to the database schema and rapid iteration is important; for complex analytics, reporting, or highly optimized queries, a hybrid approach may be preferable.