The announcement of Einstein 1 at Dreamforce this year combines Data Cloud with the new Einstein Copilot. But to make use of these enhancements, how can you make sure that you have the correct data, at the time of writing a prompt, that’s required to generate an optimal and accurate output?
Enter: Data Graphs, part of Data Cloud.
What Are Data Graphs?
When designing prompts in Prompt Builder or assigning skills in Skills Builder, you may need to check that Einstein has access to all the data required to answer a user’s prompt. This could require multiple layers of related data, going several objects deep, or data from multiple integrated sources.
How do you check that your data is ready for such exploration? Data Graphs (in Data Cloud) enable you to visualize the relationships between data model objects (DMOs), even going several layers deep. Similar to the Schema Builder, you can trace the related fields and rearrange the relationships in a drag-and-drop manner, to ensure that the correct field data will be included in an AI application.
This creates a dynamic set of data points to provide further instruction to the LLM. For example, you can ensure that Einstein has data available when surfacing outputs for prompts made by users.
How Do Data Graphs Work?
Once the Data Graph is formed (DMOs added), Data Graphs can be applied in Prompt Builder, just like a merge field. While it looks like code (because it is JSON), this is only just the context that’s been embedded in the prompt you’ve instructed.
Data Cloud Data Graphs are a way to visualize the data you have available for Einstein to answer a user’s question in Prompt Builder or assigning skills in Skills Builder. This is a nice, visual tool for us to get an understanding of our data structures.