How to Set Up AI Models in Einstein Copilot Studio

Set up and authenticate your AI models in Einstein Copilot Studio, and output your predictions in Data Cloud …

Here are highlights from article How to Set Up AI Models in Einstein Copilot Studio

1. Einstein Copilot Studio Overview
– Allows companies to train AI models with real-time customer data from Data Cloud
– Supports Bring Your Own Model (BYOM) solution
– Integrates with Salesforce Data Cloud and other AI services

2. Setting up AI Models in Einstein Copilot Studio
– Set up endpoint with name, API name, and optional description
– Configure request and response format (e.g., JSON)
– Choose authentication method (API key or JSON Web Token)

3. Configuring Model for Predictions
– Enter variables in the correct order for model predictions
– Ensuring variables are in the correct order to avoid errors

4. Outputting Inferences/Predictions
– Create or select a data model object to hold predictions
– Define fields to store predictions
– Use the appropriate naming convention for field names in JSON response format

5. Viewing Model Refresh History
– Access Refresh History tab in Einstein Copilot Studio
– View previous and current job refreshes
– Check status, date/time, and number of rows processed for each refresh

You can read it here: https://sfdc.blog/DkZVL

Source from developer(dot)salesforce(dot)com