---
title: What is Lakebase Postgres?
sidebar_label: Overview
description: Lakebase Postgres is managed Postgres inside Databricks, co-located with your Lakehouse. OLTP storage with instant branching and autoscaling.
---

# What is Lakebase Postgres?

Lakebase Postgres is managed PostgreSQL that runs inside your Databricks workspace, co-located with your workspace data and services.

Use it for the data your apps actively write and read at low latency: user state, sessions, chat history, and logs stored alongside your analytical data in the Lakehouse.

## What makes it different from running your own Postgres

- **Runs inside your workspace**, eliminating VPC peering, cross-cloud credential management, and network latency.
- **Instant branching** via copy-on-write storage creates isolated database copies in seconds, similar to git branches. Branches share unchanged data so they're cheap to create and maintain.
- **Autoscales** with your workload, scaling up under load and back down when demand drops, within a configured min/max range. No capacity planning or manual resize.
- **Scales to zero** when idle and resumes on the next query. No cost for idle compute. Development branches suspend by default after five minutes.

## How AppKit wires it up

Add the `lakebase()` plugin to `createApp` and the plugin sets up a `pg.Pool` with automatic OAuth token refresh:

```typescript
import { createApp, server, lakebase } from "@databricks/appkit";

const AppKit = await createApp({
  plugins: [server(), lakebase()],
});

// Standard pg.Pool query
const { rows } = await AppKit.lakebase.query("SELECT * FROM app.items");

// ORM-ready config (Drizzle, Prisma, etc.)
const ormConfig = AppKit.lakebase.getOrmConfig();
```

The plugin handles OAuth token refresh and connection pooling automatically. When deployed, the platform injects connection values as environment variables and the plugin reads them. No manual configuration needed. The [AppKit `lakebase` plugin reference](https://databricks.com/devhub/docs/appkit/v0/plugins/lakebase) details the pool configuration options and full API.

## When to use it

- Your app needs low-latency reads and writes: user state, sessions, conversation history, or transactional records.
- You're building AI agents that need persistent memory: conversation history, workflow state, or tool results across requests.
- You want isolated database branches for feature development or CI testing.
- You're syncing data between your OLTP workload and the [Data Lakehouse](https://databricks.com/devhub/docs/lakehouse/overview) via change data capture.

## When not to use it

- Pure analytics: read-only queries against large datasets belong in Unity Catalog, not Lakebase Postgres.
- Apps with no other Databricks workspace dependencies. The colocation advantage doesn't apply, and auth becomes your responsibility (Databricks does not inject credentials or refresh tokens for apps running outside the workspace).

## Where to next

[Templates](https://databricks.com/devhub/templates) are agent-ready prompts organized by use case. Find one that fits, or see the [Lakebase Postgres Quickstart](https://databricks.com/devhub/docs/lakebase/quickstart) for step-by-step instructions.