Pass with professional Databricks-Generative-AI-Engineer-Associate actual quiz materials

A Generative AI Engineer just deployed an LLM application at a digital marketing company that assists with answering customer service inquiries.
Which metric should they monitor for their customer service LLM application in production?

• A. Number of customer inquiries processed per unit of time
• B. Final perplexity scores for the training of the model
• C. HuggingFace Leaderboard values for the base LLM
• D. Energy usage per query

A Generative Al Engineer is working with a retail company that wants to enhance its customer experience by automatically handling common customer inquiries. They are working on an LLM-powered Al solution that should improve response times while maintaining a personalized interaction. They want to define the appropriate input and LLM task to do this.
Which input/output pair will do this?

• A. Input: Customer reviews; Output Group the reviews by users and aggregate per-user average rating, then respond
• B. Input: Customer service chat logs; Output Group the chat logs by users, followed by summarizing each user ' s interactions, then respond
• C. Input: Customer service chat logs; Output: Find the answers to similar questions and respond with a summary
• D. Input: Customer reviews: Output Classify review sentiment

A Generative Al Engineer has developed an LLM application to answer questions about internal company policies. The Generative AI Engineer must ensure that the application doesn't hallucinate or leak confidential data.
Which approach should NOT be used to mitigate hallucination or confidential data leakage?

• A. Fine-tune the model on your data, hoping it will learn what is appropriate and not
• B. Use a strong system prompt to ensure the model aligns with your needs.
• C. Add guardrails to filter outputs from the LLM before it is shown to the user
• D. Limit the data available based on the user's access level

A Generative AI Engineer is managing prompt templates using MLflow v3.x for a document summarization pipeline. A regulatory audit requires the team to demonstrate exactly which prompt version was used to generate outputs on a specific date three months ago, including the exact prompt text and any variables used at that time.
Which combination of MLflow v3.x capabilities allows the engineer to satisfy this audit requirement?

• A. MLflow Prompt Registry version history and logged runs that reference the prompt name and version used during inference.
• B. MLflow autologging and Delta Lake time travel on the inference table.
• C. MLflow Model Registry webhooks and a downstream audit log stored in an external database.
• D. MLflow experiment tags and an automatically scripted changelog stored in a Databricks notebook.

A Generative AI Engineer is building a multi-turn chat app with LangGraph on Databricks. The app must persist chat history-messages, roles, timestamps, and session IDs-for many concurrent users, support SQL queries, and stay governed in Unity Catalog. The engineer also wants ACID guarantees, low-latency reads and writes, and an easy way to sync chat data into Delta tables for analytics and model training.
Which approach fits these requirements?

• A. Write each turn from a custom LangGraph node directly into a Delta table with Spark append, then query history via Spark SQL on every request.
• B. Use a custom in-memory LangGraph state store running on the Databricks cluster driver, and periodically snapshot the state to JSON files in DBFS.
• C. Use Lakebase with a chat_history table wired to a Postgres-backed LangGraph checkpoint/memory component and enable synchronization from Lakebase into Delta tables.
• D. Store conversation history in MLflow runs and retrieve it via the MLflow Tracking API inside LangGraph nodes.