Question 1

You are developing an AI application that needs to handle and generate diverse types of content for an interactive storytelling platform. The platform must integrate text and images seamlessly, allowing users to generate stories with accompanying visuals. Which of the following approaches best leverages Multi-Model Architectures for this scenario?

Options :

A : Embed text and images into the same vector space but do not use any advanced Multi-Modal LLMs, relying on basic retrieval systems to handle each modality individually.

B : Use one modality, such as text, as the primary focus and adapt images according to predefined templates or rules, without dynamically generating or integrating images with the text.

C : Use a Multi-Modal LLM like GPT-4V(ision) to generate both the narrative text and associated images in a unified manner, allowing the system to produce a story with integrated visuals based on user input.

D : Store text and images separately in distinct databases, and use separate models for text generation and image retrieval, without integrating them into a single system.

Answer: C

Question 2

A Generative Al Engineer is responsible for developing a chatbot to enable their companys internal HelpDesk Call Center team to more quickly find related tickets and provide resolution. While creating the GenAI application work breakdown tasks for this project, they realize they need to start planning which data sources (either Unity Catalog volume or Delta table) they could choose for this application. They have collected several candidate data sources for consideration: call_rep_history: a Delta table with primary keys representative_id, call_id. This table is maintained to calculate representatives call resolution from fields call_duration and call start_time. transcript Volume: a Unity Catalog Volume of all recordings as a *.wav files, but also a text transcript as *.txt files. call_cust_history: a Delta table with primary keys customer_id, cal1_id. This table is maintained to calculate how much internal customers use the HelpDesk to make sure that the charge back model is consistent with actual service use. call_detail: a Delta table that includes a snapshot of all call details updated hourly. It includes root_cause and resolution fields, but those fields may be empty for calls that are still active. maintenance_schedule “ a Delta table that includes a listing of both HelpDesk application outages as well as planned upcoming maintenance downtimes. They need sources that could add context to best identify ticket root cause and resolution. Which TWO sources do that? (Choose two.)

Options :

A : call_cust_history

B : maintenance_schedule

C : call_rep_history

D : call_detail

E : transcript Volume

Answer: D,E

Question 3

When choosing between "Deploy Model" and "Deploy Code" patterns in a Databricks environment, which approach is generally recommended for more effective management of ML workflows?

Options :

A : Deploy Model: Move the model artifact from staging to production after training in the dev environment, and handle other code (inference, monitoring) separately.

B : Deploy Code: Develop and test code in separate environments, then deploy it to production, ensuring the training pipeline is also run in each environment while deploying the model only in production.

C : Deploy Model: Train the model in the development environment and deploy it directly to production without separate processes for inference and monitoring.

D : Deploy Code: Train the model only in the production environment and use the same environment for testing and deployment.

Answer: B

Question 4

What is a key challenge when implementing reranking in a Retrieval-Augmented Generation (RAG) pipeline for a Gen AI application?

Options :

A : Reranking simplifies the RAG pipeline by reducing the overall complexity and streamlining the process.

B : Reranking requires minimal computational resources, making it an efficient process for large datasets.

C : Reranking eliminates the need for any further processing or adjustments to the retrieved documents before generating a response.

D : Reranking involves multiple calls to the LLM, which can significantly increase both the cost and latency of the RAG chain.

Answer: D

Question 5

You are working on a recommendation system for an e-commerce platform that needs to provide personalized product suggestions based on customer behavior and preferences. The system must handle large volumes of data, including user interactions, product descriptions, and reviews. Given that your company is already using the Databricks Lakehouse platform, which approach would be most efficient for implementing this recommendation system?

Options :

A : Implement a third-party vector database and integrate it with the existing Databricks Lakehouse platform for managing and querying high-dimensional data.

B : Use a traditional relational database to store and query user data and product information, integrating it with the recommendation system.

C : Develop a custom in-house solution for vector storage and search, integrating it with the Databricks Lakehouse platform.

D : Utilize the Databricks Vector Search to store, index, and query vector embeddings of user interactions and product data, leveraging its tight integration with the Databricks Lakehouse platform.

Answer: D

Free Databricks-Certified-Generative-AI-Engineer-Associate Mock Exam – Practice Online Confidently

Buying Options: