Before you begin
You will need a SCHMA account and a customer-owned PostgreSQL 15+ database with the pgvector extension installed and enabled. Providers such as Supabase and Neon can supply a connection string. Have access to the database credentials and your provider’s connection settings.
Your content is stored in the database you connect. Your database provider’s hosting and charges are separate from your SCHMA plan. You can create and edit content manually on Free; AI tools and other Pro features depend on your workspace plan.
Open the setup options to begin.
1. Create a workspace and connect your database
Create a workspace in the Studio and follow the database onboarding flow. Enter the PostgreSQL connection URL for your PostgreSQL 15+ database (with pgvector enabled), then complete connection verification and database migrations.
Use a database intended for this workspace. A connection URL begins with postgres:// or postgresql://. It includes credentials, so keep it out of frontend code and source control.
You can review your connection under Settings → Database. If setup fails, check the database troubleshooting guide.
2. Give your content a structure
Open Content Models and use the Schema Builder to create a content type. Start with something small, such as a post with a title, a slug, and a rich text body. Add validation where it matters, and save the model.
On a workspace with AI enabled, you can describe a model to the AI assistant. Review the generated fields and adjust them to your needs. The visual builder is available when you prefer to make each change yourself.
An example post model
- Title
- Text · required
- Slug
- Slug
- Body
- Rich text
3. Write, review, and publish
Create an entry for your new content type. Fill in the fields and save a draft while you work. Review the content, then publish when it is ready for your application.
Normal delivery requests return published entries. A saved draft will not appear in your live application. See the publication checklist if your content is missing.
Create drafts with SCHMA AI
On Pro, generate staged content previews for your model, review and edit the results, then save selected items as drafts. Publishing remains a separate action.
4. Connect your frontend
Create a delivery API key in the Studio, keep it in your server environment, and configure your workspace slug. Use the JavaScript/TypeScript SDK or query the Delivery API directly.