> ## Documentation Index
> Fetch the complete documentation index at: https://oculon.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# Database types

> Understand the main database patterns in Oculon and when each one fits best.

A database can start from different ownership models. In Oculon, the most common choice is between a database you manage directly and one that is created from a query or integration workflow.

![Placeholder image for future Oculon screenshot showing manual and query-backed database creation options](https://mintcdn.com/oculon/86VwbwjFLcFRyaEK/images/databases/database-types-placeholder.svg?fit=max\&auto=format\&n=86VwbwjFLcFRyaEK\&q=85\&s=b4f0ad4860663f394e261b0826133056)

## Manual databases

Use a manual database when your team wants to define and maintain the structure directly in Oculon.

This works well for:

* planning support tables
* controlled reference lists
* operational records that do not need an external source system
* shared tables that a team updates directly in the app

## Query-backed databases

Use a query-backed database when the data should come from an external source through an integration.

This works well when:

* another system owns the records
* refresh timing matters
* the same source should feed several downstream workflows
* you want a repeatable import path instead of manual updates

In the older Oculon docs, this appears as creating a database from an integration query.

## How to choose

Choose the database type based on ownership first.

* If Oculon is the system where the team maintains the records, use a manual database.
* If another system owns the records, use a query-backed database.

That choice usually matters more than convenience because it affects refresh behavior, change control, and who is responsible for data quality.

## Mixed workflows are common

Many teams use both patterns in the same workspace.

For example:

* a query-backed database for CRM opportunity data
* a manual database for planning assumptions or mapping tables

That is often a healthy setup as long as each database has a clear owner.

## What to review before you create one

Before you choose a type, confirm:

* where the source of truth should live
* how often the records change
* who is allowed to edit the records
* whether the database will later feed models or dimensions

## Related pages

* [What is a database](https://oculon.mintlify.site/docs-new/databases/what-is-a-database.md)
* [Working with data](https://oculon.mintlify.site/docs-new/databases/working-with-data.md)
* [Refresh and sync behavior](https://oculon.mintlify.site/docs-new/databases/refresh-and-sync-behavior.md)
* [Database best practices](https://oculon.mintlify.site/docs-new/databases/database-best-practices.md)


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.