Integrating BrazeAI Decisioning Studio™

Learn how to integrate BrazeAI Decisioning Studio™ into Braze and partner with the AI Expert Services team to build agents that apply AI for 1:1 decision-making to improve your key business metrics.

Prerequisites

Before you can integrate, you’ll need an active BrazeAI Decisioning Studio™ license. Interested in learning more? Book a call.

Integrating decision studio

Step 1: Get your endpoint URL

You’ll need to get the endpoint URL associated with your specific Braze instance. For more information, see Braze API endpoints.

Step 2: Create an API key

In Braze, go to Settings > API Keys, then create a new key with the following permissions:

Permission	Purpose	Required?
/users/track	Updates custom attributes on user profiles, in addition to creating temporary user profiles when using test sends.	✓
/users/delete	Deletes temporary user profiles that were created while using test sends.	Only for test sends
/users/export/segment	Updates the available audience communications every morning by exporting the list of users from each selected segment.	✓
/users/export/ids	Retrieves a list of identifiers when targeting users using an external_id instead of a segment. Since decisioning studio doesn’t accept Personally Identifiable Information (PII), you’ll need to ensure your fields_to_export parameter returns only non-PII fields.
Only if using external_ids
/messages/send	Sends recommended variants at the recommended time using API Campaigns that are configured for decisioning studio’s experimenter.	✓
/campaigns/list	Retrieves the list of active campaigns and extracts available email content for experimentation.	✓
/campaigns/data_series	Exports aggregated campaign data to enable reporting, validation, and troubleshooting in decisioning studio, so you can compare reporting values and analyze baseline performance. While not required, this permission is recommended.
/campaigns/details	Retrieves HTML content, subject line, and image resources from existing Campaigns for experimentation.	✓
/canvas/list	Retrieves the list of active Canvases to extract available email content for experimentation.	✓
/canvas/data_series	Exports aggregated canvas data for reporting and validation, especially when BAU is orchestrated via Canvas. While not required, this permission is recommended.
/canvas/details	Retrieves HTML content, subject line, and image resources from existing Canvases for experimentation.	✓
/segments/list	Retrieves all existing segments as potential target audiences for the decisioning studio experimenter.	✓
/segments/data_series	Exports segment size information, which is shown in decisioning studio when selecting an audience.	✓
/segments/details	Retrieves segment details such as entry and exit criteria to help understand changes in audience size or performance.
/templates/email/create	Creates copies of selected base HTML templates with dynamic placeholders (Braze liquid tags) for experimentation, avoiding changes to the originals.	✓
/templates/email/update	Pushes updates to decisioning studio-created template copies when experimentation criteria change, such as call-to-actions.	✓
/templates/email/info	Retrieves information about decisioning studio-created templates in your Braze instance.	✓
/templates/email/list	Validates that templates were successfully copied over to your Braze instance.	✓

Step 3: Contact your BrazeAI Decisioning Studio™ customer success manager

Reach out to your BrazeAI Decisioning Studio™ customer success manager and ask them to enable BrazeAI Decisioning Studio™. They’ll use your Braze API key and endpoint URL to finish setting up your integration.

When it’s complete, you’ll work alongside the AI Expert Services team to start building agents for your product. Each agent is tailor-made to a specific business goal, so you’ll work together to design an implementation that’s right for you.