Databricks-GenAI-Assoc Assembling and Deploying Apps Practice Question
When evaluating an LLM for deployment, what is a crucial 'non-functional' requirement that must be addressed?
⚠ Common exam trap
Students often focus exclusively on accuracy metrics or benchmark scores, forgetting that operational factors like latency and throughput are essential non-functional requirements.
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
✓
The inference latency and throughput under expected load.
Non-functional requirements like latency, throughput, and cost are as important as the model's accuracy. In production, an accurate model that is too slow to provide a response is useless. Similarly, a high-performing model that is prohibitively expensive to run is not viable. Balancing these factors is essential for ensuring that the generative AI application is both technically feasible and financially sustainable at scale.
Answer analysis
Option-by-option breakdown
For each option: why learners choose it and why it is or isn't the right answer here.
- ✗
The model's ability to learn from user feedback automatically.
Why it's wrong here
Automatic learning from user feedback is a complex, high-risk feature that is not a standard non-functional requirement for deployment. Most production systems use a human-in-the-loop or periodic retraining workflow for feedback incorporation to avoid data poisoning or unpredictable behavior that can arise from autonomous, real-time model updates.
- ✓
The inference latency and throughput under expected load.
Why this is correct
Latency and throughput are critical non-functional requirements for production AI applications. If an application cannot respond within a timeframe acceptable to the user or handle the expected volume of concurrent requests, it will fail to meet business objectives, regardless of how accurate or intelligent the model is.
- ✗
The number of parameters in the model architecture.
Why it's wrong here
The number of parameters is a proxy metric for model size, but it is not a direct non-functional requirement for deployment. The focus should be on the performance characteristics like latency and memory consumption, not just the parameter count, as some models may be optimized to run efficiently despite size.
- ✗
The language in which the model was initially trained.
Why it's wrong here
While language training is important for model capability, it is a functional property rather than a non-functional requirement for deployment. The deployment phase is primarily concerned with operational metrics, whereas functional capability is handled during the testing and evaluation phases that precede the deployment decision.
About these practice questions
This Databricks-GenAI-Assoc question is part of Courseiva's 330-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. 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.