Helping Hand Questions for Salesforce-AI-Specialist

Thestandard Einstein Copilot actionthat would be used in response to the user utterance, “Show me all the customers in New York,” isQuery Records. This action is responsible for retrieving a set of records from Salesforce based on a specified condition — in this case, filtering customers by location (New York).

Query Recordsis the action that fetches relevant data based on the criteria provided in the user's input.

Select Recordsis more about picking specific records from an already presented list.

Fetch Recordsis not a standard term used in this context for the action.

Refer toEinstein Copilot documentationon how Copilot actions work with natural language queries and data retrieval.

When implementingEinstein Generative AIfor improved customer insights and interactions, theData Cloudis a key consideration for storing and managing large-scale audit and feedback data. TheSalesforce Data Cloud(formerly known asCustomer 360 Audiences) is designed to handle and unify massive datasets from various sources, making it ideal for storing data required for AI-powered insights and reporting. By provisioningData Cloud, organizations likeUniversal Containers (UC)can gain real-time access to customer data, making it a central repository for unified reporting across various systems.

Audit and feedback datagenerated by Einstein Generative AI needs to be stored in a scalable and accessible environment, and theData Cloudprovides this capability, ensuring that data can be easily accessed for reporting, analytics, and further model improvement.

Custom objectsorSalesforce Big Objectsare not designed for the scale or the specific type of real-time, unified data processing required in such AI-driven interactions.Big Objectsare more suited for archival data, whereasData Cloudensures more robust processing, segmentation, and analysis capabilities.

References:

Salesforce Data Cloud Documentation:

Salesforce Einstein AI Overview:

To implement a solution where users enter a sales order number and the system generates a summary, the AI Specialist should create ascreen flowto collect the sales order number and invoke the prompt template. The standard"Prompt Template" flow actioncan then be used to trigger the custom prompt, providing a summary of the sales order header and details.

Option B, creating a template-triggered prompt flow, is not necessary for this scenario because the requirement is to directly collect input through a screen flow.

Option C, using an autolaunched flow, would be inappropriate here because the solution requires user interaction (entering a sales order number), which is best suited to a screen flow.

Salesforce AI Specialist References:For further guidance on creating prompt templates with flows:

Universal Containersshould use aPredictive modelto decide which of the three emails to send based on the customer'slifetime value scoreandmarket segment. Predictive models analyze data to forecast outcomes, and in this case, it would predict the most appropriate email to send based on customer attributes. Additionally, predictive models can provideexplainabilityto show why a certain email was chosen, which is crucial for UC’s requirement to explain the decision-making process.

Generative modelsare typically used for content creation, not decision-making, and thus wouldn't be suitable for this requirement.

Predictive modelsoffer the ability to explain why a particular decision was made, which aligns with UC’s needs.

Refer toSalesforce’s Predictive AI model documentationfor more insights on how predictive models are used for segmentation and decision making.