Question 1
Multiple choice
You are implementing Agentic AI within an Enterprise AI Factory. You are focused on the operation and scaling of the agentic systems including each of the Enterprise AI Factory components. Which observability strategy involves providing detailed insights into the system's performance? (Choose two.)
-
A
Detailed model and application tracing for identifying performance bottlenecks.
-
B
Centralized logging to track system events.
-
C
Continuous monitoring of key metrics using OpenTelemetry (OTEL).
-
D
Artifact repository used by the AI agents where all the system performance metrics are stored.
Reveal answer details
Close answer details
Correct answersA, C
ExplanationDetailed model and application tracing reveals where latency, failures, or bottlenecks occur inside the agentic system. Continuous monitoring of key metrics with OpenTelemetry keeps performance signals visible over time, which is the core observability need for operating at scale.
You're working with an LLM to automatically summarize research papers. The summaries often omit critical findings. What's the best way to ensure that the summaries accurately reflect the core insights of the research papers?
-
A
Asking the LLM to "summarize the paper."
-
B
Asking the LLM to "understand" the paper to generate a summary.
-
C
Having the LLM generate the summaries and then manually review every output.
-
D
Asking the LLM to "extract the key findings."
Reveal answer details
Close answer details
Correct answerD
ExplanationAsking the LLM to extract the key findings directs it toward the specific information the summaries were missing. A broad summarize instruction can compress the paper without preserving the core research insights.
A development team is building a customer support agent that interacts with users via chat. The agent must reliably fetch information from external databases, handle occasional API failures without crashing, and improve its responses by learning from user feedback over time. Which of the following tasks is most critical when enhancing an AI agent to handle real-world interactions and improve over time?
-
A
Applying a well-structured training process with foundational generative models and prompt engineering
-
B
Utilizing internal knowledge bases to support agent responses alongside external APIs
-
C
Implementing retry logic for error handling and integrating user feedback loops for iterative improvement
-
D
Designing conversation flows that provide consistent responses based on predefined scripts
Reveal answer details
Close answer details
Correct answerC
ExplanationReal-world support agents need error handling and a way to improve. Retry logic prevents occasional API failures from crashing the interaction, while user feedback loops support iterative improvement in future responses.
A financial services agentic AI is being used to automate initial customer onboarding. The agent is completing the process efficiently and accurately, but reviews of its conversations reveal it often uses overly formal and complex language that confuses customers. Which type of evaluation is best suited to address this issue?
-
A
Controlled user testing sessions to collect user feedback on the clarity and tone of responses
-
B
Compliance review of the agent's access to regulatory guidelines and policy documentation
-
C
Continuous user feedback collection, specifically gathering subjective assessments of the agent's communication style
-
D
Statistical analysis of the agent's decision-making patterns to detect overly formal and complex response choices
Reveal answer details
Close answer details
Correct answerA
ExplanationOverly formal or complex language is best evaluated through controlled user testing because real users can judge clarity and tone directly. Compliance review or processing statistics would not capture whether customers understand the responses.
When analyzing inconsistent performance across a fleet of customer service agents handling similar queries, which evaluation approach most effectively identifies root causes and optimization opportunities?
-
A
Assess performance data from recently improved agents and highlight strong results, using outcome comparisons to identify areas with the greatest impact on service quality.
-
B
Average performance metrics across all agents as this will smooth individual variations, query distribution differences, and temporal factors affecting agent behavior and accuracy.
-
C
Deploy stratified evaluation sampling across agent variants, query complexity levels, and temporal patterns while tracking decision paths using comparative analytics.
-
D
Review performance across both high-and low-accuracy agent groups, comparing case outcomes and identifying patterns contributing to top and bottom results.
Reveal answer details
Close answer details
Correct answerC
ExplanationStratified evaluation sampling separates results by agent variant, query complexity, and time pattern, which helps reveal root causes of inconsistent service. Comparative analytics on decision paths can show which conditions lead to weaker outcomes.
You're developing an agent that monitors social media mentions of your brand. The social media platform's API returns data mentioning your brand with varying confidence scores that the brand was actually being mentioned, but these scores aren't consistently calibrated. Considering the unreliability of these confidence scores, what's the most reliable way for the agent to insure it is truly processing media mentions of the brand?
-
A
Using an approach that filters mentions with basic keyword search and removes those with exceptionally low confidence scores, relying on the API data as a first-pass filter.
-
B
Using an approach that treats all mentions as equally reliable, regardless of their confidence scores, and applies a uniform data processing workflow to minimize inconsistency.
-
C
Using a threshold-based approach, accepting mentions only if their confidence score exceeds a predefined level that aligns with typical thresholds used for well-calibrated APIs.
-
D
Using an approach that combines the agent's text analysis with the API's confidence score, weighing the agent's assessment more heavily when identifying mentions.
Reveal answer details
Close answer details
Correct answerD
ExplanationUncalibrated confidence scores should not be trusted as the sole filter. Combining the agent's text analysis with the API score, while weighing the agent assessment more heavily, gives the system a direct check on whether the brand is truly being mentioned.
When analyzing a customer service agentic system's performance degradation over time, which evaluation approach most effectively identifies opportunities for human-in-the-loop intervention to improve agent decision-making transparency and user trust?
-
A
Monitor only final task completion rates without examining intermediate decision points, user interaction patterns, or opportunities for beneficial human intervention during agent conversations
-
B
Implement multi-stage evaluation tracking decision confidence scores, user correction patterns,
intervention effectiveness, and explainability-satisfaction correlations
-
C
Rely on periodic manual reviews of random conversation samples without systematic tracking of intervention effectiveness, decision transparency, or user trust indicators
-
D
Collect anonymous usage statistics without capturing specific decision rationales, user feedback on agent explanations, or transparency improvement opportunities for trust building
Reveal answer details
Close answer details
Correct answerB
ExplanationPerformance degradation tied to trust needs more than completion counts. Multi-stage evaluation of confidence scores, user correction patterns, intervention effectiveness, and explainability satisfaction shows where human review improves decisions.
When evaluating optimization opportunities between NeMo Guardrails, NIM microservices, and TensorRT- LLM in a production healthcare agent, which analysis approach best identifies optimization opportunities across the NVIDIA stack?
-
A
Conduct stress testing of individual microservices and guardrails to measure peak throughput and determine theoretical performance limits of each module.
-
B
Use default configurations to establish a deployment baseline, focusing on stability before conducting deeper performance profiling.
-
C
Create end-to-end latency waterfalls that capture guardrail overhead, NIM queuing delays, and TensorRT optimization benefits while assessing overall pipeline efficiency.
-
D
Tune each component individually, focusing primarily on local performance metrics with secondary attention to integration patterns.
Reveal answer details
Close answer details
Correct answerC
ExplanationEnd-to-end latency waterfalls expose where time is spent across guardrail overhead, NIM queuing, and TensorRT optimization effects. Looking at the whole pipeline identifies integration bottlenecks that isolated component tuning can miss.
An agent is tasked with solving a series of complex mathematical problems that require external tools to find information. It often struggles to keep track of intermediate steps and reasoning. Which prompting technique would be MOST effective in improving the agent's clarity and reducing errors in its reasoning?
-
A
-
B
-
C
-
D
Reveal answer details
Close answer details
Correct answerA
ExplanationReAct interleaves reasoning with tool use, so the agent can decide what to do, call an external tool, observe the result, and continue reasoning from that evidence. That structure is well suited to complex math tasks where intermediate steps are easy to lose.
Question 10
Single choice
A development team is building an AI agent capable of autonomously planning and executing multi-step tasks while retaining context and learning from past interactions. 66 66 Which practice is most important to enable the agent to effectively manage long-term memory and complex tasks?
-
A
Implement memory mechanisms for context retention and apply chain-of-thought prompts to enhance reasoning.
-
B
Use basic rule-based decision methods that emphasize fast responses over adaptive planning.
-
C
Apply short-term memory approaches that handle each interaction independently of previous ones.
-
D
Reduce planning features and memory management to keep the system streamlined.
Reveal answer details
Close answer details
Correct answerA
ExplanationAutonomous multi-step work depends on remembering relevant context and structuring reasoning across steps. Memory mechanisms support context retention, while chain-of-thought prompting helps the agent break down decisions instead of treating each interaction as isolated.
Question 11
Multiple choice
When evaluating a multi-agent customer service system experiencing unpredictable scaling costs and performance bottlenecks during peak hours, which analysis approaches effectively identify optimization opportunities for both infrastructure efficiency and service reliability? (Choose two.)
-
A
Maintain consistent resource allocation across all service hours, for a more precise view of baseline traffic impact on long-term infrastructure efficiency.
-
B
Scale agent infrastructure based on aggregate performance trends, using system-wide monitoring tools to identify broader optimization patterns across resources.
-
C
Deploy agents with configurable scaling workflows, allowing analysis of resource adjustment strategies and their effects on service stability during variable demand periods.
-
D
Deploy distributed tracing with cost attribution per agent type, correlating resource consumption with business value metrics to identify optimization opportunities in agent deployment strategies.
-
E
Implement comprehensive workload profiling using NVIDIA Nsight to analyze GPU utilization patterns, identify underutilized resources, and optimize batch sizing for dynamic scaling with Kubernetes HPA.
Reveal answer details
Close answer details
Correct answersD, E
ExplanationDistributed tracing with cost attribution separates what each agent type consumes and ties that cost to business value, which exposes inefficient deployment patterns. Workload profiling with NVIDIA Nsight then reveals GPU utilization, idle capacity, and batch-sizing issues that affect scaling and service reliability under peak demand.
Question 12
Single choice
When designing complex agentic workflows that include both sequential and parallel task execution, which orchestration pattern offers the greatest flexibility?
-
A
Graph-based workflow orchestration incorporating conditional branches
-
B
Linear pipeline orchestration with a fixed task sequence
-
C
Event-driven orchestration that triggers tasks reactively, in series or in parallel
Reveal answer details
Close answer details
Correct answerA
ExplanationGraph-based workflow orchestration is the most flexible pattern because a graph can model sequential paths, parallel branches, joins, and conditional routing. A fixed linear pipeline cannot represent that variety without becoming brittle.
Question 13
Single choice
Your agent is designed to manage tasks through a service management API. The API responds with detailed event logs, but these logs contain both metadata and structured data. To ensure the agent correctly interprets and processes the data from these logs, what's the most prudent approach?
-
A
Employ a specialized parser that adheres to the API's documentation, to insure strict adherence to structured data.
-
B
Employing a modular design that allows the agent to dynamically adjust its parsing logic.
-
C
Using a human-in-the-loop approach, manually inspecting and interpreting each log entry.
-
D
Employ a specialized parser that extracts all data fields, regardless of their type.
Reveal answer details
Close answer details
Correct answerA
ExplanationA parser built to follow the API documentation is the prudent choice because structured logs must be interpreted according to the provider's defined fields, types, and formats. Dynamic or field-grabbing approaches increase the chance of treating metadata as usable structured data.
Question 14
Single choice
Optimize agentic workflow performance with the NVIDIA Agent Intelligence Toolkit.
Your organization is building a complex multi-agent system that needs to connect agents built on different frameworks while maintaining optimal performance. 66 66 Which key features of the NVIDIA Agent Intelligence Toolkit would be MOST beneficial for this implementation?
-
A
The toolkit is limited to simple agent-to-agent communication but cannot orchestrate complex multi- agent workflows.
-
B
The toolkit provides framework-agnostic integration ensuring reusability of components.
-
C
The toolkit is designed exclusively for NVIDIA framework agents and cannot integrate with other frameworks.
-
D
The toolkit focuses primarily on agent development but lacks evaluation capabilities.
Reveal answer details
Close answer details
Correct answerB
ExplanationFramework-agnostic integration is the useful feature when agents come from different frameworks. It allows components to be reused and orchestrated across the system instead of locking the implementation to one agent framework or only simple communication.
Question 15
Single choice
When implementing stateful orchestration for agentic workflows using LangGraph, which memory management approach provides the best balance of performance and context retention?
-
A
Store complete conversation history in memory with periodic database syncing
-
B
Implement rolling window memory with fixed conversation length limits
-
C
Use session-ID based checkpointer with user-defined schema for selective state persistence
Reveal answer details
Close answer details
Correct answerC
ExplanationLangGraph stateful workflows need durable state without carrying every token forever. A session-ID based checkpointer stores the right state per session, while a user-defined schema controls what persists, balancing retrieval speed with useful context retention.
Question 16
Single choice
Which memory design best supports a subscription renewal assistant that must remember customer preferences across sessions while still responding quickly during the current chat?
-
A
A fixed window containing only the most recent chat turns.
-
B
Hybrid memory with short-term state plus long-term vector retrieval and key-value records.
-
C
Periodic model retraining as the only way to store customer history.
-
D
Sending the full lifetime conversation history to the model on every turn.
Reveal answer details
Close answer details
Correct answerB
ExplanationA renewal assistant needs fast short-term state for the active chat and durable memory for customer preferences across sessions. Hybrid memory combines that immediate context with long-term vector retrieval and key-value records, avoiding both a tiny recent window and costly full-history prompts.
Question 17
Multiple choice
To launch a customer support pilot that must learn from resolved chats and survive intermittent billing API timeouts, which two design choices are appropriate? Choose two.
-
A
Replace API calls with canned answers for common account questions.
-
B
Capture user feedback and resolution outcomes for iterative behavior updates.
-
C
Make a static decision tree the primary conversation manager.
-
D
Add bounded retry handling for transient external API failures.
-
E
Remove external integrations until the pilot has stable traffic volume.
Reveal answer details
Close answer details
Correct answersB, D
ExplanationA support pilot improves by capturing user feedback and resolution outcomes, then using them for iterative behavior updates. Bounded retry handling addresses transient billing API failures without replacing real integrations or making failures unbounded.
Question 18
Single choice
You are designing an AI agent for summarizing medical documents that include images and text as well. It must extract key information and recognize dates. Which feature is most critical for ensuring the agent performs well across multiple input and output formats?
-
A
Use of guardrails to filter out hallucinated content
-
B
Retry logic implementation to ensure robustness during API failures
-
C
Chain-of-thought prompting for reasoning accuracy
-
D
Multi-modal model integration to handle both text and vision inputs
Reveal answer details
Close answer details
Correct answerD
ExplanationA medical summarization agent handling documents with images and text needs multi-modal model integration. That capability lets the system process both vision inputs and text inputs, which is more central here than retries or guardrails alone.
Question 19
Single choice
When analyzing safety violations in a financial advisory agent that uses NeMo Guardrails, which evaluation approach best identifies gaps in guardrail coverage?
-
A
Apply keyword-and rule-based validation methods to confirm compliance with policy terms and common risk conditions.
-
B
Analyze violation patterns, test adversarial prompts, measure guardrail activation, and align policies with observed failures.
-
C
Conduct functional testing with representative user inputs to verify policy enforcement in typical usage scenarios.
-
D
Monitor overall guardrail activations and system logs to assess operational behavior across different interaction types.
Reveal answer details
Close answer details
Correct answerB
ExplanationCoverage gaps are found by looking beyond ordinary functional tests. Analyzing violation patterns, testing adversarial prompts, measuring guardrail activation, and aligning policies with observed failures shows where financial-safety rules fail to trigger or need refinement.
Question 20
Single choice
A Lead AI Architect at a global financial institution is designing a multi-agent fraud detection system using an agentic AI framework. The system must operate in real time, with distinct agents working collaboratively to monitor and analyze transactional patterns across accounts, retain and share contextual information over time, and escalate suspicious behaviors to a human fraud analyst when needed. Which architectural approach enables intelligent specialization, shared memory, and inter-agent coordination in a dynamic and evolving threat environment?
-
A
Design a modular multi-agent system where individual agents collaborate asynchronously using shared memory and structured messaging.
-
B
Design a multi-agent system where individual agents collaborate synchronously using shared memory and structured messaging.
-
C
Design a centralized rule-based service that checks all transactions against static fraud indicators and sends alerts when thresholds are exceeded.
-
D
Design an agentic workflow where each agent acts independently on isolated data slices with no inter-agent communication to reduce latency and model complexity.
-
E
Design monolithic LLM-based agents that handle all fraud detection tasks within a single loop, without modular roles or multi-agent coordination.
Reveal answer details
Close answer details
Correct answerA
ExplanationFraud detection across accounts benefits from modular agents with specialized roles, shared memory, and structured messages. Asynchronous collaboration supports real-time work because agents can monitor, analyze, and escalate without waiting on a single lockstep sequence.
Question 21
Single choice
Enable what in a document drafting assistant so attorneys can review AI-suggested clauses before filing and improve future recommendations?
-
A
Inline explanations, cited risk factors, accept or modify controls, and immediate feedback capture.
-
B
A batch accept-all control followed by a summary report after review.
-
C
Automatic insertion of every suggested clause with human review only at the end.
-
D
A hidden recommendation score that is stored for later developer analysis.
Reveal answer details
Close answer details
Correct answerA
ExplanationAttorney review needs visibility and control before filing. Inline explanations and cited risk factors make the suggested clause reviewable, accept or modify controls keep the human decision in the workflow, and immediate feedback capture helps improve future recommendations.
Question 22
Single choice
You are deploying a multi-agent customer-support system on Kubernetes using NVIDIA GPU nodes and Triton Inference Server. Traffic spikes during product launches. You need <100ms response times, zero
downtime, automatic GPU scaling, and full monitoring. Which deployment setup best achieves cost-effective, reliable, low-latency scaling?
-
A
Set up one mixed GPU node pool with Cluster Autoscaler min=0, scale by network throughput, monitor via metrics-server and logs, and skip readiness probes for fast startup.
-
B
Place GPU pods on on-demand nodes in one zone, disable Cluster Autoscaler, run a fixed pod count for bursts, scale on CPU usage, and monitor with default health checks.
-
C
Deploy GPU pods in a node pool spanning all zones, mix GPU types, enable Cluster and Horizontal Pod Autoscalers using Prometheus GPU and latency metrics, and monitor with NVIDIA DCGM and Grafana.
-
D
Use spot-instance node pools across zones, enable Cluster Autoscaler with capped nodes, scale on memory usage, and monitor with logs and cluster events.
Reveal answer details
Close answer details
Correct answerC
ExplanationSub-100 ms service during launch bursts needs scaling tied to GPU pressure and latency, not just CPU or memory. A node pool spanning zones improves availability, Cluster and Horizontal Pod Autoscalers add capacity, and Prometheus with DCGM and Grafana gives the monitoring needed to tune reliability.
|