Databricks-GenAI-Assoc Assembling and Deploying Apps Practice Question
An engineer deployed a RAG agent to a Model Serving endpoint and enabled inference tables for payload logging. After a week, reviewers notice that logged requests contain customer email addresses and order identifiers. Compliance requires that raw prompts and responses not be stored in plain text. Which change should the engineer make?
⚠ Common exam trap
The trap here is treating inference table logging as all-or-nothing, when Unity Catalog masking and grants let you keep logs while redacting sensitive fields.
Answer choices
Why each option matters
Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.
Correct answer & explanation
✓
Keep inference tables enabled but restrict SELECT on the inference table to a compliance group and apply row filters or column masks on the sensitive columns.
Because inference tables are Unity Catalog tables, the engineer can apply column masks and row filters to redact or restrict sensitive fields and grant SELECT only to authorized principals. This preserves the observability inference tables provide while preventing plain-text exposure of personal data, which is the precise compliance outcome required.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
Enable payload logging but write inference tables to a workspace-local path outside Unity Catalog.
Why it's wrong here
Moving the table outside Unity Catalog removes the governance controls that make masking and auditing possible, and it does not redact anything. The sensitive values would still be stored in plain text, now with weaker access control. This worsens the compliance posture rather than addressing it.
- ✗
Disable inference tables entirely so no request or response data is persisted.
Why it's wrong here
Disabling inference tables eliminates the sensitive payload logging, but it also removes observability that the team relies on for monitoring quality and debugging. Compliance requires protecting the data, not abandoning logging. A targeted redaction or access-control approach preserves usefulness while meeting the privacy requirement.
- ✗
Configure the endpoint to log only the model's token counts and latency, then reconstruct prompts from the application's own logs.
Why it's wrong here
Logging only metrics avoids storing payloads but shifts the sensitive data problem to the application logs, which are not governed by the endpoint. Reconstructing prompts elsewhere does not satisfy a requirement that raw prompts not be persisted, and it fragments observability across systems.
- ✓
Keep inference tables enabled but restrict SELECT on the inference table to a compliance group and apply row filters or column masks on the sensitive columns.
Why this is correct
Inference tables are Unity Catalog tables, so column masks and row filters can redact or hide sensitive fields, and GRANT controls who can query them. This keeps observability for authorized reviewers while ensuring raw personal data is not broadly readable, satisfying the compliance requirement without losing monitoring capability.
About these practice questions
One of 330 original Databricks-GenAI-Assoc practice questions on Courseiva, each with a full explanation and wrong-answer analysis — not exam dumps or protected exam content. Learn why practice questions differ from exam dumps →
JA
Written and reviewed by Johnson Ajibi, MSc IT Security
Senior Network & Security Engineer · founder of Courseiva
Last reviewed September 2026 · checked against the official Databricks exam blueprint
This Databricks-GenAI-Assoc practice question is part of Courseiva's free Databricks certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the Databricks-GenAI-Assoc exam.