Deploy RDS

Lexia provides a built-in PostgreSQL database for AI agent developers. This removes the need to set up, host, or manage your own database infrastructure. Using a single CLI command, Lexia automatically creates a fully configured RDS instance and returns all required connection details.

The database is ready to use immediately and can be connected directly to your agent code for long-term or short-term memory storage.

1. Overview

Lexia's Relational Database provides a managed PostgreSQL environment that eliminates the need for manual database setup, hosting, or configuration. With the Lexia CLI, you can create, manage, and connect to PostgreSQL databases with minimal effort.

What this enables:

Once provisioned, your agent can immediately:

Store conversation memory
Save user data
Manage long-term storage
Handle custom application logic
Persist any structured information

This database is fully managed by Lexia, so you don't need to deploy, maintain, or scale it yourself.

2. How it Works

When you run the database creation command, Lexia provisions a PostgreSQL database for your workspace.

The returned information includes:

Database name
Schema
Username
Password
Workspace details
Connection string

No Docker setup, no manual deployment, and no hosting configuration are required.

You can focus entirely on your agent instead of database operations.

3. Install CLI

Install globally via npm:

npm install -g @lexia/cli

Or run without installation:

npx @lexia/cli kickstart python

4. Log in

Authenticate with your workspace:

lexia login

Select Team Mode, enter your workspace name, and provide your authentication token.

The token can be generated under API Keys in the Lexia dashboard.

5. Create Database

List available DB commands:

lexia db

Create a new database:

lexia db create

Lexia will generate a PostgreSQL instance and return the credentials.

These include schema, username, password, and a ready-to-use connection string.

You can insert these details directly into your agent's code or environment variables.

Example Output

After running lexia db create, you'll receive output similar to:

Database created successfully!

Database Name: lexia_db_abc123
Schema: public
Username: lexia_user
Password: secure_password_here
Connection String: postgresql://lexia_user:[email protected]:5432/lexia_db_abc123

6. Change Password

To change the database password:

lexia db password

You'll be prompted to select the database and enter a new password. The connection string will be updated automatically.

Alternatively, you can specify the database and new password directly:

lexia db password --database <database-name> --new-password <new-password>

7. Connect to Database

Once you have the connection details from lexia db create, you can connect to your database in several ways:

Using Environment Variables

Add the connection string to your environment:

export LEXIA_DB_URL="postgresql://lexia_user:[email protected]:5432/lexia_db_abc123"

Using Connection String Directly

Python example:

import psycopg2

conn = psycopg2.connect(
    host="db.lexiaplatform.com",
    database="lexia_db_abc123",
    user="lexia_user",
    password="your_password",
    port=5432
)

Node.js example:

const { Client } = require('pg');

const client = new Client({
    host: 'db.lexiaplatform.com',
    database: 'lexia_db_abc123',
    user: 'lexia_user',
    password: 'your_password',
    port: 5432
});

client.connect();

Using SDK Configuration

Python:

from lexia import lexia

lexia.config(database_url="postgresql://lexia_user:[email protected]:5432/lexia_db_abc123")

Node.js:

const lexia = require('@lexia/sdk');

lexia.config({ 
    databaseUrl: "postgresql://lexia_user:[email protected]:5432/lexia_db_abc123" 
});

8. Delete Database

To delete a database:

lexia db delete

You'll be prompted to select the database to delete. Confirm the deletion when prompted.

Warning: This action cannot be undone. All data in the database will be permanently lost.

Alternatively, you can specify the database name directly:

lexia db delete --database <database-name>