What is PostgreSQL

What it is

Postgres is a relational database that stores data in typed tables with enforced constraints and relationships, queried with SQL. Its extension system means it can also do document storage (JSONB), full-text search and vector similarity without bolting on another system.

• Full ACID transactions and strong data integrity
• JSONB, full-text search, PostGIS, and pgvector extensions
• Rock-solid, standards-compliant SQL

When we reach for it

When data is relational and integrity matters — orders, accounts, billing, anything where a half-written transaction is unacceptable — Postgres is our default. Its JSONB support also lets it flex when part of the model is genuinely schemaless.

Trade-offs

Relational schemas require upfront design and disciplined migrations; changing them later takes care. Horizontal scaling across many machines is more involved than with some NoSQL stores. For fast-changing, deeply nested, document-shaped data, MongoDB can be the better fit.

Why it matters

PostgreSQL ("Postgres") is a powerful, open-source relational database renowned for reliability, correctness and standards compliance, and it has become the default choice for serious applications that need a robust, trustworthy data store. It matters because data integrity is non-negotiable for most businesses, and Postgres delivers rock-solid ACID transactions and strong consistency while also being remarkably feature-rich — supporting advanced data types, full-text search, JSON, and extensibility that let it handle workloads people once reached for specialist databases to do. Being open-source and free, with no vendor lock-in, adds to its appeal.

• A powerful, open-source relational database known for reliability and correctness
• The default for applications needing robust, trustworthy data storage
• Feature-rich and extensible, free and open-source with no vendor lock-in

Key characteristics

Postgres is a relational database using SQL, with full ACID compliance (atomicity, consistency, isolation, durability) that guarantees transactions are processed reliably — critical wherever data correctness matters, like finance or orders. Beyond the relational basics it is unusually capable: native JSON/JSONB support (so it can handle document-style data), full-text search, geospatial data (via PostGIS), a rich type system, and extensibility through extensions. Its trade-off versus lighter or specialist databases is that its power comes with more to configure and tune for scale, but for the vast majority of applications it is a dependable, feature-complete default.

• Relational and SQL-based with full ACID compliance for reliable transactions
• Strong data integrity and consistency — critical for correctness-sensitive data
• Feature-rich: JSON/JSONB, full-text search, geospatial, extensibility
• More to tune at scale than lightweight databases, but dependable and complete

When to use it

PostgreSQL is an excellent default for almost any application that needs a reliable relational store — transactional systems, SaaS back-ends, analytics, anything where data integrity and correctness matter. Its JSON support means it can also handle semi-structured data, often removing the need for a separate document database. You might reach elsewhere for specialised needs — a document database for schemaless flexibility at scale, an in-memory store for caching, a specialist analytics engine for massive OLAP — but even then Postgres frequently suffices. The pragmatic default for most teams is "use Postgres unless you have a specific reason not to".

• An excellent default for transactional systems and SaaS back-ends
• Its JSON support often removes the need for a separate document database
• Specialised needs may warrant a document, in-memory or analytics engine
• Common wisdom: use Postgres unless you have a specific reason not to