Databricks Generative AI Engineer Associate exam dumps

Databricks Generative AI Engineer Associate practice question 188 of 306

Databricks Certified Generative AI Engineer Associate. Free level, Databricks. Free question with the correct answer and a full explanation.

Databricks Generative AI Engineer Associate Question 188

Select 3

You are setting up a Databricks model serving endpoint for a machine learning model that provides predictions for a financial fraud detection system. It is critical to ensure that only authorized applications can access the endpoint. Which of the following steps should you take to control access to the model serving endpoint?

  1. A

    Use access tokens to authenticate requests to the model serving endpoint.

  2. B

    Enable IP allowlisting for the model serving endpoint.

  3. C

    Disable all authentication mechanisms to allow public access for testing purposes.

  4. D

    Integrate the endpoint with a permissions model using Databricks workspace access control.

  5. E

    Leverage SSL/TLS encryption to secure the communication channel.

Show answer and explanation

Correct answers: A, B, D

Explanation

To control access to model serving endpoints in Databricks, you should use access tokens for authentication, enable IP allowlisting to restrict access to trusted IPs, and integrate with workspace access control to enforce fine-grained permissions. These measures ensure that only authorized users and applications can access the endpoint while securing the system against unauthorized access. While SSL/TLS is important for securing communication, it does not directly control endpoint access.

  • A. Correct.

    Using access tokens is a recommended approach for authenticating requests to Databricks resources, including model serving endpoints, ensuring only authorized users or applications can access the endpoint.

  • B. Correct.

    IP allowlisting can restrict access to specific IP addresses, providing an additional layer of security for the model serving endpoint.

  • C. Incorrect.

    Disabling authentication mechanisms for public access is highly insecure and should never be done in a production setup, especially for sensitive applications like financial fraud detection.

  • D. Correct.

    Integrating the endpoint with Databricks workspace access control allows you to enforce fine-grained access permissions, ensuring only authorized users and groups can interact with the endpoint.

  • E. Incorrect.

    While SSL/TLS encryption is critical for securing data in transit, it does not control access to the endpoint but rather secures the communication channel.

Timed practice exam

Take a Databricks Generative AI Engineer Associate practice test under exam conditions

45 questions in 90 minutes, drawn from this bank, with a score report and a per-question review when you finish.

Start timed exam