You are here:
Einstein Report Formula Generation Considerations and Limitations
Before working with Einstein Report Formula Generation, understand the considerations and limitations.
Required Editions
| Available in: Lightning Experience. |
| Available in: Enterprise, Performance, Unlimited, and Developer Editions with Einstein 1 Sales Edition and Einstein 1 Service Edition. |
- Generative AI can produce inaccurate or harmful responses. Always review the output for accuracy and safety. You assume responsibility for how Einstein’s responses are applied to your company.
- Einstein Report Formula Generation isn’t conversational, and follow-up questions aren’t supported. If Einstein’s response isn’t what you want, try again by rephrasing your request.
- Your requests and Einstein’s responses in the Einstein panel aren’t retained. When you close the formula builder, the requests and responses aren’t saved.
- When editing a formula, if you click Apply, the new formula overwrites the old one.
- Einstein generative AI uses a Large Language Model (LLM) to generate the text and formulas upon your request. Sometimes, the LLM response isn’t what you expect. If this happens, you can generate another response by starting over.
- Report Formula Creation uses your data to add context and customization to your request. The Report Formula Generation feature references the data behind the report you’re building. Your data stays inside Salesforce and isn’t shared with the LLM. Your data is used only to generate relevant outputs, and it trains only models that are hosted by Salesforce. No public data is used.
- The formulas suggested by Einstein are subject to the limitations of reports. For a complete list of limitations, see these pages.
See Also
Did this article solve your issue?
Let us know so we can improve!
