Select Chain Components for a Desired Model Input and Output GENERATIVE-AI-ENGINEER-ASSOCIATE Practice Questions (Page 2)

Questions 6–10

1. 6expert · hard

   A team is building a chain that answers questions by first retrieving relevant documents, then generating an answer, and finally checking the answer for factual consistency against the retrieved documents. The consistency check uses a separate model. The team wants to minimize latency. Which orchestration approach is most efficient?

   • ARun the retrieval and consistency check in parallel, then generate the answer
   • BRun generation and retrieval in parallel, then run the consistency check
   • CRun the consistency check first, then retrieval, then generation
   • DRun retrieval, then generate the answer, then run the consistency check sequentially
   Reveal Solution Discussion Select an answer first
2. 7application · medium

   A team is building a chain that uses a tool to look up current stock prices. The tool expects a ticker symbol as input, but users type company names like 'Databricks Inc.' in a free-text field. The chain must convert the user's input into the format the tool requires. Which component should be added to the chain to handle this conversion?

   • AA prompt template that instructs the model to extract the company name and return the ticker symbol
   • BAn output parser that converts the model's response into a JSON object
   • CA retriever that searches a database of company information
   • DA memory component that stores the user's previous queries
   Reveal Solution Discussion Select an answer first
4. 8application · easy

   A team is building a chain that classifies customer support tickets into categories (e.g., 'billing', 'technical', 'account'). The downstream system expects the category as a single string. The model sometimes returns extra text like 'The category is billing.' The team needs to ensure the output is exactly 'billing'. Which component should be added?

   • AAn output parser that extracts the category from the model's response
   • BA memory component to store the classification result
   • CA prompt template that asks for a one-word answer
   • DA retriever to fetch similar tickets for reference
   Reveal Solution Discussion Select an answer first
5. 9application · easy

   A developer is building a chatbot that answers follow-up questions about a user's previous requests. The chain must remember what the user asked earlier in the conversation to provide contextually relevant answers. Which chain component should be included to achieve this?

   • AA prompt template that includes the user's current question only
   • BA memory component that stores and injects conversation history into the prompt
   • CAn output parser that formats the model's responses as chat messages
   • DA retriever that fetches documents from a knowledge base
   Reveal Solution Discussion Select an answer first
6. 10foundation · easy

   A chain requires the model to receive the user's question and a list of retrieved documents. What is the correct way to format this input for the model?

   • AProvide the documents only in the system message and the question in the user message
   • BCombine the question and documents into a single prompt string with clear delimiters
   • CSend the question and documents as separate API calls to the model
   • DPass the documents as a separate binary attachment
   Reveal Solution Discussion Select an answer first