Databricks Certified Generative AI Engineer Associate · Free Practice Question Hard
Question 27
A Generative AI Engineer must create a pipeline that retrieves financial reports and summarizes key performance metrics. The system must handle thousands of reports efficiently. What is the correct sequence for designing this pipeline?
- A 1. Deploy the retriever → 2. Store raw reports in a database → 3. Summarize retrieved reports → 4. Index the database.
- B 1. Preprocess all reports → 2. Deploy a retriever → 3. Fine-tune a language model → 4. Summarize outputs.
- C 1. Fine-tune a summarization model → 2. Generate embeddings → 3. Build a retriever → 4. Deploy to production.
- D 1. Create embeddings for all financial reports → 2. Store embeddings in a Vector Search index → 3. Build a retriever → 4. Deploy a summarization model for retrieval outputs.
Reveal correct answer
Correct answer: D
Explanation
Embedding generation, vector indexing, retrieval, and summarization ensure the pipeline is scalable and optimized for financial document processing.A. Incorrect: Indexing must occur before retrieval for efficient querying.
B. Incorrect: Preprocessing and retrieval must include indexing for effective querying.
C. Incorrect: Summarization models are applied after retrieval, not before embedding generation.
D. Correct: This sequence ensures that documents are indexed for retrieval before being summarized, optimizing both retrieval and generation processes.
Discussion
Think the marked answer is wrong, or have a better explanation? Share it below — comments appear after review.
You must be logged in to post a comment.
