What criteria do you use when choosing a database management system for a new project?

When choosing a database management system (DBMS) for a new project, I consider the following criteria:

• Data type and structure: relational (SQL) or non-relational (NoSQL). If data is strictly structured and requires complex relationships — I choose SQL; if flexible schemas or large volumes — NoSQL.

• Scalability: horizontal or vertical scalability, support for clusters.

• Performance: read/write speed, latency, transaction requirements.

• Transactional consistency: whether strict ACID compliance is needed or if eventual consistency suffices.

• Ecosystem and support: availability of drivers, tools, community.

• Ease of integration with the technology stack used (e.g., Node.js).

• Licensing and cost.

• Team experience with a specific DBMS.

For example, for a project with a large amount of unstructured data and high load, MongoDB is often chosen, while for financial applications with critical transactions — PostgreSQL or MySQL.