AI-201 · Question #143
A data science team has trained an XGBoost classification model for product recommendations on Databricks. The Agentforce Specialist is tasked with bringing inferences for product recommendations…
The correct answer is A. Create the serving endpoint in Databricks, then configure the model using Model Builder. When integrating a Databricks-trained model (like XGBoost) into Salesforce Data Cloud, the workflow is: (1) Create a serving endpoint in Databricks - this exposes the model as a REST API endpoint that Salesforce can call. (2) Configure the connection in Salesforce using Model…
Question
A data science team has trained an XGBoost classification model for product recommendations on Databricks. The Agentforce Specialist is tasked with bringing inferences for product recommendations from this model into Data Cloud as a stand-alone data model object (DMO). How should the Agentforce Specialist set this up?
Options
- ACreate the serving endpoint in Databricks, then configure the model using Model Builder.
- BCreate the serving endpoint in Einstein Studio, then configure the model using Model Builder.
- CCreate the serving endpoint in Databricks, then configure the model using a Python SDK
How the community answered
(36 responses)- A75% (27)
- B17% (6)
- C8% (3)
Explanation
When integrating a Databricks-trained model (like XGBoost) into Salesforce Data Cloud, the workflow is: (1) Create a serving endpoint in Databricks - this exposes the model as a REST API endpoint that Salesforce can call. (2) Configure the connection in Salesforce using Model Builder (in Einstein Studio), which registers the external endpoint and maps its outputs to a stand-alone DMO in Data Cloud. Option B is incorrect because serving endpoints are created in Databricks, not Einstein Studio. Option C is incorrect because the Python SDK is not the mechanism for bringing model inferences into Data Cloud as a DMO.
Topics
Community Discussion
No community discussion yet for this question.