Databricks-GenAI-Assoc Design Applications Practice Question
A GenAI engineer is designing a multi-turn chat application on Databricks. Users report that the assistant forgets details from earlier in long conversations and sometimes answers using only the most recent message. The team wants the model to reliably use facts stated several turns earlier without exceeding the model's context window. Which design should the engineer implement?
⚠ Common exam trap
The trap here is assuming the model has inherent memory across API calls, when in fact each inference request is stateless and only sees the prompt it is given.
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
✓
Implement a conversation memory strategy that summarizes or selectively retains earlier turns and injects the compressed history into each prompt within the context window.
Reliable multi-turn memory requires bringing earlier relevant turns into the prompt, either by summarizing them or by selectively retaining key facts. This keeps the payload within the context window while preserving details the user expects the assistant to remember. Temperature changes, sending only the latest message, or storing transcripts without retrieval all fail to put earlier facts in front of the model.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✓
Implement a conversation memory strategy that summarizes or selectively retains earlier turns and injects the compressed history into each prompt within the context window.
Why this is correct
A memory strategy that summarizes or selectively retains earlier turns keeps essential facts available while staying within the context window. By injecting the compressed history into each prompt, the model receives the earlier details it needs, which directly addresses the reported forgetting and avoids the cost and truncation problems of sending the full transcript.
- ✗
Store the full conversation transcript in a Delta table and instruct the model to query the table during inference.
Why it's wrong here
A chat model served through Model Serving cannot autonomously query a Delta table during generation; it only sees the prompt it receives. Without a retrieval step that fetches relevant rows and injects them into the prompt, storing the transcript in Delta does not make earlier facts available to the model and does not solve the context window issue.
- ✗
Increase the model's temperature so it explores more of the conversation history when generating a response.
Why it's wrong here
Temperature controls randomness in token sampling, not how much of the conversation the model attends to. Raising it makes responses less deterministic and more prone to drift, but it does not cause the model to retrieve or weight earlier turns, so the memory problem remains and answer quality may degrade.
- ✗
Send only the latest user message to the model and rely on the model's pretrained knowledge to recall earlier facts.
Why it's wrong here
The model has no persistent memory across API calls and cannot recall facts from earlier turns unless they are included in the prompt. Sending only the latest message guarantees that earlier details are unavailable, which is precisely the failure users are reporting, and it does not address the context window constraint.
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.