SQLAlchemy Python ORM

SQLAlchemy is the Python SQL toolkit and Object Relational Mapper that gives application developers the full power and flexibility of SQL. Created by Michael Bayer and first released in 2006, the library has become the most widely used database abstraction layer in the Python ecosystem, powering major applications at companies including Reddit, Yelp, Quora, and Dropbox. The current version 2.x features a comprehensive rewrite with full type annotation support, async/await compatibility, and modern Python 3 patterns. The architecture consists of two distinct layers: SQLAlchemy Core provides a schema centric SQL expression language with type coercion, expression compilation, and connection pooling that generates SQL across 20 plus database backends including PostgreSQL, MySQL, SQLite, Oracle, Microsoft SQL Server, and MariaDB. SQLAlchemy ORM builds on Core to provide a high level data mapper pattern with support for classical mappers, declarative class definitions, and the modern annotated declarative style. Key ORM features include eager and lazy loading strategies with joined, subquery, and selectin loading, relationship management with backreferences and cascade rules, versioned transactions with session scoping and identity maps, hybrid properties that work in both Python and SQL contexts, and polymorphic mappings for inheritance hierarchies. The engine provides connection pooling with configurable pool size, timeout, and recycle settings, and supports savepoints, two phase commits, and isolation level control. The MIT licensed project has over 10,000 GitHub stars.