You are here:
Considerations for Return Insights
To create Return Insights with Einstein Generative AI, consider supported functionality, usage, limitations and allowances, limits, and other issues.
Required Editions
| View supported editions. |
Return Insights Language and Locale Support
Return Insights supports these languages.
| LOCALE | CODE |
|---|---|
| English (United Kingdom) | en_UK |
| English (United States) | en_US |
Return Insights Large Language Model Support
Return Insights supports these models.
| MODEL |
|---|
| OpenAI GPT-4 |
Einstein Trust Layer Service Support
Return Insights supports these Trust Layer services.
| SERVICE | DESCRIPTION |
|---|---|
| Dynamic Grounding with Secure Data Retrieval | Full support. Return Insights uses an internal, standard prompt that passes through a list of preselected returned products. |
| Prompt Defense (System policies and prompt injection detection) | Full support. |
| Data Masking | Full support. |
| Toxicity Detection | Full support. |
| Audit and Feedback | There’s no explicit user feedback component for users, but feedback is sent to a feedback service via an API. |
Ask your Salesforce admin about which Einstein Trust Layer services are enabled in your org and available for Return Insights.
Return Insights Region Support
Return Insights supports these models for geo-aware request routing.
| MODEL |
|---|
| sfdc_ai__DefaultGPT4 |
Return Insights and Einstein Requests Usage
Einstein Requests is a usage metric for generative AI. The use of generative AI capabilities, in either a production or sandbox environment, consumes Einstein Requests and possibly Data 360 credits. To learn more, see Einstein Usage.
Other Considerations
Return Insights integrates with all Order Management supported integrations if you have the required licenses and permissions.
