I'm building a product that needs JSON fields and relational data, what database fits best?

If you need both JSON fields and relational data, the default best fit for most products is PostgreSQL.

Why PostgreSQL

• Strong relational support: tables, joins, constraints, transactions, foreign keys.
• Excellent JSON support: JSON / JSONB columns let you store flexible data without losing SQL querying.
• Hybrid model: you can keep core data relational and use JSON for semi-structured or optional fields.
• Mature ecosystem: indexing, ORMs, backups, replication, analytics, and broad tooling.

When PostgreSQL is especially a good choice

• Your schema is mostly structured, but some fields vary by record.
• You need to query inside JSON occasionally.
• You want consistency and transactional safety.
• You expect the product to evolve and don’t want to redesign the schema constantly.

Alternatives

• MySQL: workable, but JSON/querying/indexing is generally less flexible than PostgreSQL.
• MongoDB: great for document data, but weaker if relational integrity and complex joins are important.
• SQLite: good for local/small apps, but not ideal for a multi-user production backend.
• CockroachDB / YugabyteDB: if you need distributed SQL, but they add complexity.

Practical recommendation

Use:

• Relational tables for core entities and relationships
• JSONB for optional/custom/variable attributes
• Proper indexes on both relational columns and JSON paths you query often

Rule of thumb

• If you need joins + transactions + some JSON → PostgreSQL
• If you need mostly flexible document storage → MongoDB
• If you want simple and safe default for mixed data → PostgreSQL

If you want, I can also help you choose based on your expected scale, query patterns, and whether you need multi-tenancy or search.