Testing Week 3 · Saturday
Day 20: Full Timed Practice Exam + Weak Area Drilling
Learning Objectives
- - Simulate real exam conditions with timed 180-minute practice exam
- - Identify remaining weak domains after 3 weeks of study
- - Drill domain-specific questions for weakest areas
Tasks
Tasks
0/2 completedHands-On Lab
Build real muscle memory with these activities.
intermediate 30 min
Build the Production RAG Pipeline Reference Architecture
Trace the full Production RAG Pipeline architecture on paper and identify all service interactions.
- 1 Draw the architecture: User → API Gateway → Lambda → Bedrock KB (Retrieve) → OpenSearch Serverless
- 2 Add the generation path: Lambda → Bedrock InvokeModel (with retrieved context) → Response
- 3 Add guardrails: Guardrail applied on both input and output
- 4 Add monitoring: CloudWatch Logs, X-Ray tracing, model invocation logging
- 5 Label every IAM role and permission needed for each service interaction
intermediate 30 min
Build the Multi-Agent System Reference Architecture
Trace the Multi-Agent System architecture and identify the orchestration patterns.
- 1 Draw the architecture: User → Supervisor Agent → Routing Decision
- 2 Add collaborator agents: Research Agent (KB tool), Action Agent (Lambda tools), Analysis Agent (Code Interpreter)
- 3 Add AgentCore components: Gateway for routing, Memory for context, Policy for access control
- 4 Add the observability layer: AgentCore Observability with OTEL spans
- 5 Note the exam-critical distinction: supervisor mode (one agent delegates) vs supervisor-with-routing mode (dynamic selection)
Practice Questions
34 questions across 3 difficulty levels.
Further Reading
Go deeper into today's topics.
GenAI Application Builder Architecture Overview
End-to-end reference architecture for production GenAI applications.
docs 30 min
Multi-Agent Orchestration with Reasoning
Supervisor mode vs supervisor-with-routing mode for multi-agent systems.
blog 25 min
Serverless Generative AI Architectural Patterns
API Gateway + Lambda + Bedrock foundational patterns for production deployments.
blog 25 min
Well-Architected GenAI Lens
Architecture patterns and best practices for GenAI workloads.
docs 45 min