You are building a customer-support chatbot that fetches user account data from an external billing API.
During testing, the API sometimes returns timeouts or 500 errors. You want the agent to be resilient-retrying when appropriate but failing gracefully if the service is down.
Which strategy best handles intermittent failures in API calls while still ensuring a good user experience?

• A. Schedule retries using a fixed delay for all failure types, maintaining predictable timing and user notifications after each attempt.
• B. Retry requests with a consistent short delay after each failure and notify the user as each retry takes place.
• C. Return a standard fallback message on failures to maintain conversation flow and reduce the risk of service interruptions for the user.
• D. Implement exponential-backoff retries with a circuit breaker, and return a clear message to the user if all retries fail.

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 monolithic LLM-based agents that handle all fraud detection tasks within a single loop, without modular roles or multi-agent coordination.
• B. Design a modular multi-agent system where individual agents collaborate asynchronously using shared memory and structured messaging.
• C. Design an agentic workflow where each agent acts independently on isolated data slices with no inter- agent communication to reduce latency and model complexity.
• D. Design a multi-agent system where individual agents collaborate synchronously using shared memory and structured messaging.
• E. Design a centralized rule-based service that checks all transactions against static fraud indicators and sends alerts when thresholds are exceeded.

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. Compliance review of the agent's access to regulatory guidelines and policy documentation
• B. Statistical analysis of the agent's decision-making patterns to detect overly formal and complex response choices
• C. Controlled user testing sessions to collect user feedback on the clarity and tone of responses
• D. Continuous user feedback collection, specifically gathering subjective assessments of the agent's communication style

A company is deploying a multi-agent AI system to handle large-scale customer interactions. They want to ensure the system is highly available, cost-effective, and scalable across multiple NVIDIA GPUs using container orchestration tools.
Which practice is most crucial for successfully deploying and scaling an agentic AI system in production?

• A. Deploy agents on a single machine to obtain a dimensioning baseline and thereby reduce setup complexity before expanding system scope.
• B. Optimize GPU utilization frameworks with workload optimization separate from cost analysis, prioritizing resource performance for peak load scenarios in deployment.
• C. Use a static assignment of requests across agents to maintain consistent agent operation and simplify coordination while scaling infrastructure resources as needed.
• D. Implementing automated workload management and resource scheduling frameworks to optimize GPU utilization and maintain service availability.