Getting started with GibsonAI

This tutorial walks you through creating your first database schema, exploring your project visually, deploying it, and accessing your auto-generated APIs.

1. Sign Up

   1. Visit https://app.gibsonai.com/
   2. Click Sign up
   3. You can use your Google account or email to create an account
   4. Once you’re in, you’ll be taken to the GibsonAI app dashboard

   ---

2. Create a New Project with a Prompt

   GibsonAI uses prompts to help you define your database schema without writing SQL.

   1. Type your schema idea in plain English. For example:

      “Create a schema with users, destinations, bookings, and reviews. Each user can make bookings and leave reviews. Each destination has a name, description, price, and rating.”

   2. Click Generate Schema

      GibsonAI will create your schema instantly using the prompt you entered.

   ---

3. Explore the Schema Diagram

   Once your schema is generated, click the Diagram tab.

   • You’ll see all your tables, relationships, and keys laid out visually with Entity-Relationship Diagram (ERD)
   • This helps you understand how your schema is structured
   • Click on any table to view details and relationships

   ---

4. Deploy Your Schema to a Database

   Now it's time to deploy your schema to a real database.

   1. Click the Deploy button (top right)
   2. Choose whether to deploy to Development or Production
   3. If you’re on the Free plan, only Production is available by default
   4. Gibson will provision a real serverless database, provides connection details and generates ready to use Data API for CRUD operations.

   Currently, GibsonAI supports MySQL database deployment. You can also connect to your existing MySQL database to import existing schemas.

   ---

5. View Your Auto-Generated API Docs

   After deployment, click the API docs tab.

   Here, you’ll see:

   • All your endpoints
   • Query parameters, response formats, and example payloads
   • Ready-to-use endpoints for testing and integrating with your app
   • Playground to test API requests

   Now test your API directly in the API Docs playground, using cURL commands or connect it to Postman. Also, you can copy code snippets for different programming languages to add into your frontend or backend code, or integrate with low-code solutions like Retool. No need to write REST controllers for typical CRUD operations.

6. Make a Schema Change

   GibsonAI supports schema evolution with automatic migrations.

   Let’s say you want to add an address column to the user table.

   1. Open the prompt bar

   2. Type:

      “Add an address column to the user table”

   3. GibsonAI updates the schema and shows the diff

   Gibson will automatically:

   • Generate the migration
   • Update your schema history
   • Keep your API updated

   Then redeploy your project to apply the changes in your database and live API.