Portal Customization

Customize branding, welcome screens, and workspace layouts for your portal

Preview: AI Builder Portal is in preview. Features may change as the product evolves.

Overview

Make the AI Builder Portal your own with custom branding, a tailored welcome screen, and per-blueprint layout configuration. These settings let you align the portal with your company identity and control the development experience your builders get.

Branding

Navigate to Admin > Settings to configure portal-wide branding.

Setting	Description
Portal Name	Displayed in the header and browser tab title. Use your company name or a team-specific label.
Logo (Light)	Your company logo for light mode, shown in the top navigation bar. Upload a PNG or SVG file.
Logo (Dark)	Your company logo for dark mode. If not provided, the light logo is used in both modes. Upload a PNG or SVG file.
Accent Color	The portal generates a full color palette from your accent color - buttons, links, accents, and hover states all adapt automatically. Use the color picker or enter a hex value.

Use the Reset controls next to the accent color and logos to revert to the Qovery defaults at any time.

Branding changes take effect immediately for all users. No restart or redeployment is required. The portal automatically applies the appropriate logo variant based on the user’s system theme (light or dark mode).

Welcome Screen

Configure the welcome screen that builders see when they first log in or when they have no active workspaces. Navigate to Admin > Settings > Welcome Screen.

Setting	Description
Title	The headline on the welcome screen (e.g., “Welcome to Acme Dev Portal”).
Description	A short paragraph explaining what the portal is and how to get started.
CTA Button Text	The label on the call-to-action button (e.g., “Create Your First Workspace”).

Use the welcome screen to onboard new builders. Explain what they can do in the portal, which blueprints are available, and where to get help. A clear welcome screen reduces support requests and gets builders productive faster.

Layout Configuration

Each blueprint can have its own layout settings that control what builders see inside the workspace editor. Navigate to Admin > Blueprints > [Blueprint] > Workspace.

Tabs

Choose which tabs appear in the workspace editor. Available tab types:

• Chat - The OpenCode chat panel for AI-assisted development
• Claude Code - Terminal tab with Claude Code pre-configured
• Terminal - A standard terminal session in the workspace container
• Custom command tabs - Tabs that run a specific command when opened

Custom Tabs

Define custom tabs that execute a specific command in the workspace container when the tab is opened. This is useful for common development tasks.

Examples:

• A tab named “Dev Server” that runs npm run dev
• A tab named “Django Server” that runs python manage.py runserver
• A tab named “Tests” that runs npm test -- --watch

Custom tabs give builders quick access to frequently used commands without needing to remember or type them.

Lock Layout

Enable Lock Layout to prevent builders from adding or removing tabs in the workspace editor. When locked, builders see exactly the tabs you configured - no more, no less.

Use this when you want a consistent, controlled experience across all workspaces using this blueprint.

Preview Panel

Toggle the preview panel visibility for the blueprint. When enabled, builders see a live preview of their running application directly in the workspace editor with viewport switching for desktop, tablet, and mobile.

Disable the preview panel for blueprints where a live preview is not relevant (e.g., backend API services, CLI tools, data processing scripts).

AI & Data Tab

The AI & Data tab combines AI provider configuration and a data sources reference grid into a single view. Navigate to Admin > Blueprints > [Blueprint] > AI & Data.

Data Sources

The AI & Data tab includes a Data Sources grid showing supported database types. Each tile represents a database or data warehouse that can be connected to your blueprint:

PostgreSQL

MySQL

Snowflake

BigQuery

Click any tile to view setup instructions for connecting that data source to your blueprint environment. The instructions guide you through configuring the necessary environment variables and connection credentials in your Qovery blueprint environment.

Data source connections are configured at the Qovery environment level using environment variables. The Data Sources grid provides setup guidance for each database type - the actual credentials and connection parameters are managed through your blueprint’s Qovery environment configuration.

AI Configuration

The AI configuration uses an accordion interface where each provider can be independently enabled and configured.

Anthropic (Claude)

Controls both Claude Code (terminal) and can power the Chat panel.

Setting	Description
Enabled	Toggle to enable or disable Anthropic as a provider for this blueprint.
Authentication Mode	Choose between User OAuth (builders authenticate with their own Anthropic account) or API Key (you provide a key, billed to your organization).
API Key	Required when using API Key authentication mode. Stored encrypted with AES-256-GCM.
Model	The Claude model to use (e.g., claude-sonnet-4-20250514).

OpenAI

Powers the Chat panel with OpenAI models.

Setting	Description
Enabled	Toggle to enable or disable OpenAI as a provider.
API Key	Your OpenAI API key. Stored encrypted with AES-256-GCM.
Model	The model to use (e.g., gpt-4o).

Custom Provider

Use any OpenAI-compatible API endpoint as the Chat panel provider.

Setting	Description
Enabled	Toggle to enable or disable the custom provider.
API Key	API key for the custom endpoint. Stored encrypted with AES-256-GCM.
Model	The model identifier expected by the custom endpoint.
Endpoint URL	The base URL of the OpenAI-compatible API (e.g., https://your-llm-gateway.internal/v1).

You can enable multiple providers simultaneously. Anthropic powers Claude Code, while the Chat panel uses whichever provider is configured for it (Anthropic, OpenAI, or Custom). If multiple chat providers are enabled, the first enabled provider in order (Anthropic, OpenAI, Custom) is used.

For organizations that require AI traffic to stay within your network, use the Custom provider with an internally hosted model endpoint. This ensures no AI data leaves your infrastructure.

Start with the default layout and iterate based on builder feedback. You can change layout settings at any time - they apply to new workspaces and existing workspaces on their next load.

Next Steps

Blueprint Management

Register and configure blueprint templates.

Access Control

Control who can access each blueprint.

Member Management

Invite and manage team members.