| Exam Name | AIP-C01 Practice Exam – AWS Certified Generative AI Developer Professional (2026 Updated) |
|---|---|
| Exam Provider | Amazon Web Services (AWS) |
| Certification Type | Professional-Level Certification (Generative AI, LLM Applications, RAG, Prompt Engineering & AI Architecture) |
| Total Practice Questions | 150 Advanced MCQs (Scenario-Based + RAG + Bedrock + Vector DB + Prompt Engineering + Security) |
| Exam Domains Covered | • Generative AI Fundamentals (LLMs, tokens, embeddings, inference) • AWS AI Services (Amazon Bedrock, SageMaker, AI APIs) • Prompt Engineering & Optimization (temperature, tokens, guardrails) • Retrieval-Augmented Generation (RAG) & Vector Databases • Model Customization (fine-tuning, embeddings, evaluation) • Security & Responsible AI (guardrails, moderation, data privacy) • Performance Optimization (latency, caching, cost control) • Monitoring & Observability (CloudWatch, logging, evaluation pipelines) |
| Questions in Real Exam | • Total: ~75 Questions • Highly scenario-driven with real-world GenAI use cases • Focus on architecture decisions, optimization, and safety |
| Exam Duration | • Total Time: 180 Minutes • Complex multi-step scenarios requiring deep reasoning • Emphasis on applied AI system design and optimization |
| Passing Score | • Scaled Score: 750 / 1000 • Requires strong understanding of GenAI concepts and AWS integration • Focus on real-world problem-solving and architecture trade-offs |
| Question Format | • Multiple Choice & Multiple Response • Scenario-Based GenAI Application Design • RAG Pipelines & Vector Search Questions • Prompt Engineering & Optimization Cases • Security, Guardrails & Responsible AI Scenarios |
| Difficulty Level | Advanced to Expert (Professional-Level + Real-World GenAI Scenarios) |
| Key Knowledge Areas | • Amazon Bedrock (foundation models, inference APIs, embeddings) • RAG architecture (chunking, retrieval, re-ranking, hybrid search) • Prompt engineering (temperature, top-k, grounding instructions) • Vector databases (ANN indexing, similarity search, optimization) • Model customization (fine-tuning vs RAG trade-offs) • Security (IAM, encryption, guardrails, prompt injection prevention) • Monitoring (CloudWatch, evaluation datasets, hallucination tracking) • Cost & latency optimization (token usage, caching, model selection) |
| Common Exam Traps | • Overusing fine-tuning instead of RAG • Ignoring prompt injection and security risks • Choosing large models when smaller ones suffice • Poor chunking strategies leading to irrelevant retrieval • Not using re-ranking or hybrid search in RAG pipelines • Ignoring token cost and latency optimization • Missing evaluation and monitoring strategies • Lack of guardrails for safe AI outputs |
| Skills Developed | • Designing production-grade GenAI applications • Building scalable RAG pipelines with vector databases • Optimizing prompts for accuracy, cost, and performance • Implementing AI safety, guardrails, and compliance • Monitoring and evaluating LLM outputs effectively • Architecting multi-agent and event-driven AI workflows |
| Study Strategy | • Focus on RAG architecture and retrieval optimization • Practice prompt engineering and temperature tuning • Learn Bedrock APIs and model selection strategies • Understand embeddings, vector search, and indexing • Study guardrails, moderation, and AI security risks • Analyze real-world GenAI scenarios and trade-offs • Take full-length timed mock exams • Review explanations to identify hidden exam traps |
| Best For | • AI/ML engineers building LLM-based applications • Software developers working with generative AI • Cloud engineers implementing AI pipelines on AWS • Professionals transitioning into GenAI and LLM engineering roles |
| Career Benefits | • Validates advanced Generative AI and LLM development skills • Opens roles in AI engineering, ML engineering, and GenAI architecture • Enhances expertise in RAG, prompt engineering, and AI pipelines • Increases earning potential in AI-driven industries • Positions you as a specialist in next-generation cloud AI solutions |
| Updated | 2026 Latest Version – Based on AWS AIP-C01 Exam Guide & Real GenAI Architecture Patterns |
1.
A developer wants to build an LLM-powered app without managing models. What is BEST?
A. SageMaker training
B. Amazon Bedrock
C. EC2
D. RDS
Answer: B
Rationale: Amazon Bedrock provides managed access to foundation models without requiring infrastructure management, enabling rapid development of generative AI applications.
2.
A developer wants to generate text responses using a foundation model. What is BEST?
A. S3
B. Bedrock InvokeModel API
C. EC2
D. DynamoDB
Answer: B
Rationale: The InvokeModel API allows developers to send prompts to foundation models and receive generated outputs in real time.
3.
A developer wants to store embeddings for semantic search. What is BEST?
A. RDS
B. Vector database
C. S3
D. EC2
Answer: B
Rationale: Vector databases store embeddings for similarity search, enabling semantic retrieval in AI applications.
4.
A developer wants to improve LLM responses using external data. What is BEST?
A. Fine-tuning only
B. Retrieval-Augmented Generation (RAG)
C. EC2
D. S3
Answer: B
Rationale: RAG combines LLMs with external knowledge sources, improving accuracy and reducing hallucinations.
5.
A developer wants to generate embeddings. What is BEST?
A. Bedrock embedding model
B. S3
C. EC2
D. RDS
Answer: A
Rationale: Bedrock provides embedding models that convert text into vectors for semantic tasks.
6.
A developer wants to reduce hallucinations in LLM output. What is BEST?
A. Increase tokens
B. Use RAG
C. Use EC2
D. Use S3
Answer: B
Rationale: RAG grounds responses in real data, reducing hallucinations and improving reliability.
7.
A developer wants to manage prompts effectively. What is BEST?
A. Hardcode prompts
B. Prompt templates and versioning
C. EC2
D. S3
Answer: B
Rationale: Prompt templates ensure consistency and allow iteration and optimization.
8.
A developer wants to fine-tune a model. What is BEST?
A. Bedrock fine-tuning
B. S3
C. EC2
D. DynamoDB
Answer: A
Rationale: Bedrock supports fine-tuning to customize models for specific use cases.
9.
A developer wants to monitor model performance. What is BEST?
A. CloudTrail
B. CloudWatch
C. Config
D. Lambda
Answer: B
Rationale: CloudWatch tracks metrics and logs for monitoring AI applications.
10.
A developer wants to secure model access. What is BEST?
A. Public access
B. IAM policies
C. EC2
D. S3
Answer: B
Rationale: IAM controls access securely.
11.
A developer wants real-time inference. What is BEST?
A. Batch processing
B. Bedrock API
C. EC2
D. S3
Answer: B
Rationale: Bedrock supports real-time inference.
12.
A developer wants to store training data. What is BEST?
A. S3
B. EC2
C. RDS
D. DynamoDB
Answer: A
Rationale: S3 is ideal for storing datasets.
13.
A developer wants chatbot functionality. What is BEST?
A. Bedrock + Lambda
B. EC2
C. S3
D. RDS
Answer: A
Rationale: Combining Bedrock with Lambda enables serverless chatbot logic.
14.
A developer wants scalable AI APIs. What is BEST?
A. API Gateway + Lambda
B. EC2
C. S3
D. RDS
Answer: A
Rationale: API Gateway and Lambda scale automatically.
15.
A developer wants to log AI responses. What is BEST?
A. CloudWatch Logs
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: Logs help debugging.
16.
A developer wants to secure data. What is BEST?
A. IAM
B. KMS
C. CloudWatch
D. Lambda
Answer: B
Rationale: KMS encrypts data.
17.
A developer wants to reduce latency. What is BEST?
A. Increase tokens
B. Use caching
C. EC2
D. S3
Answer: B
Rationale: Caching reduces repeated requests.
18.
A developer wants event-driven AI workflows. What is BEST?
A. EventBridge
B. EC2
C. RDS
D. S3
Answer: A
Rationale: EventBridge triggers workflows.
19.
A developer wants batch inference. What is BEST?
A. Real-time API
B. Batch processing with SageMaker
C. EC2
D. S3
Answer: B
Rationale: SageMaker supports batch inference.
20.
A developer wants AI pipeline automation. What is BEST?
A. Step Functions
B. EC2
C. S3
D. RDS
Answer: A
Rationale: Step Functions orchestrate workflows.
21.
A developer wants vector similarity search. What is BEST?
A. RDS
B. Vector DB
C. S3
D. EC2
Answer: B
Rationale: Vector DB enables similarity search.
22.
A developer wants prompt optimization. What is BEST?
A. Trial and error
B. Prompt engineering techniques
C. EC2
D. S3
Answer: B
Rationale: Prompt engineering improves results.
23.
A developer wants scalable storage. What is BEST?
A. S3
B. EC2
C. RDS
D. DynamoDB
Answer: A
Rationale: S3 scales automatically.
24.
A developer wants authentication. What is BEST?
A. IAM
B. Cognito
C. S3
D. EC2
Answer: B
Rationale: Cognito manages user auth.
25.
A developer wants monitoring alerts. What is BEST?
A. CloudWatch alarms
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: Alarms notify issues.
26.
A developer wants API security. What is BEST?
A. IAM
B. API Gateway authorizer
C. S3
D. EC2
Answer: B
Rationale: Authorizers secure APIs.
27.
A developer wants data transformation. What is BEST?
A. Lambda
B. EC2
C. S3
D. RDS
Answer: A
Rationale: Lambda processes data.
28.
A developer wants high availability. What is BEST?
A. Single AZ
B. Multi-AZ
C. EC2
D. S3
Answer: B
Rationale: Multi-AZ ensures redundancy.
29.
A developer wants cost optimization. What is BEST?
A. Use EC2
B. Use serverless
C. Use RDS
D. Use S3
Answer: B
Rationale: Serverless reduces cost.
30.
A developer wants scalable AI apps. What is BEST?
A. Lambda + Bedrock
B. EC2
C. S3
D. RDS
Answer: A
Rationale: Serverless AI architecture scales automatically.
31.
A RAG system returns irrelevant results due to poor retrieval quality. What is BEST?
A. Increase tokens
B. Improve embedding model and chunking strategy
C. Use EC2
D. Use S3
Answer: B
Rationale: Retrieval quality depends heavily on embedding accuracy and document chunking. Better embeddings and optimized chunk sizes improve semantic matching and relevance of retrieved results.
32.
A developer wants to reduce hallucinations in a chatbot. What is BEST?
A. Increase temperature
B. Use RAG with verified data sources
C. Use EC2
D. Use S3
Answer: B
Rationale: Grounding responses with trusted data using RAG significantly reduces hallucinations and improves factual correctness.
33.
A developer wants consistent responses from an LLM. What is BEST?
A. High temperature
B. Low temperature
C. Use EC2
D. Use S3
Answer: B
Rationale: Lower temperature reduces randomness and produces more deterministic outputs, improving consistency for production systems.
34.
A developer wants to protect against prompt injection attacks. What is BEST?
A. Ignore
B. Input validation and guardrails
C. Use EC2
D. Use S3
Answer: B
Rationale: Prompt injection can manipulate LLM behavior. Input validation, guardrails, and filtering ensure safe and controlled outputs.
35.
A developer needs fast semantic search at scale. What is BEST?
A. RDS
B. Vector database with indexing
C. S3
D. EC2
Answer: B
Rationale: Vector databases use indexing (e.g., ANN) to enable fast similarity searches across large embedding datasets.
36.
A developer wants to reduce inference latency. What is BEST?
A. Increase tokens
B. Use smaller model or caching
C. Use EC2
D. Use S3
Answer: B
Rationale: Smaller models and caching reduce response time and improve user experience without sacrificing performance significantly.
37.
A developer wants to evaluate LLM output quality. What is BEST?
A. Ignore
B. Automated evaluation metrics + human review
C. EC2
D. S3
Answer: B
Rationale: Combining automated metrics with human evaluation ensures comprehensive assessment of model performance.
38.
A developer wants to version prompts. What is BEST?
A. Hardcode
B. Prompt versioning system
C. EC2
D. S3
Answer: B
Rationale: Versioning allows tracking changes, rollback, and continuous improvement of prompts in production.
39.
A developer wants scalable embeddings generation. What is BEST?
A. Manual
B. Batch processing with Bedrock or SageMaker
C. EC2
D. S3
Answer: B
Rationale: Batch processing efficiently generates embeddings at scale for large datasets.
40.
A developer wants to reduce cost of LLM usage. What is BEST?
A. Increase tokens
B. Optimize prompts and use caching
C. Use EC2
D. Use S3
Answer: B
Rationale: Prompt optimization reduces token usage, and caching avoids repeated inference calls, lowering costs.
41.
A developer wants multi-turn conversation memory. What is BEST?
A. Ignore history
B. Store conversation context externally
C. EC2
D. S3
Answer: B
Rationale: Maintaining conversation history externally allows context-aware responses and better user experience.
42.
A developer wants secure API access. What is BEST?
A. Public
B. API Gateway + IAM/Cognito
C. S3
D. EC2
Answer: B
Rationale: API Gateway with authentication ensures secure access to AI services.
43.
A developer wants real-time streaming responses. What is BEST?
A. Batch processing
B. Streaming APIs
C. EC2
D. S3
Answer: B
Rationale: Streaming APIs provide token-by-token responses, improving user experience in chat applications.
44.
A developer wants to detect toxic content. What is BEST?
A. Ignore
B. Content moderation model
C. EC2
D. S3
Answer: B
Rationale: Moderation models filter harmful content and ensure compliance with safety guidelines.
45.
A developer wants data privacy. What is BEST?
A. Public access
B. Encryption + access control
C. EC2
D. S3
Answer: B
Rationale: Encryption and IAM policies protect sensitive data.
46.
A developer wants to fine-tune models efficiently. What is BEST?
A. Train from scratch
B. Fine-tuning with domain data
C. EC2
D. S3
Answer: B
Rationale: Fine-tuning improves model performance for specific use cases without full retraining.
47.
A developer wants to optimize retrieval speed. What is BEST?
A. Linear search
B. Vector indexing (ANN)
C. EC2
D. S3
Answer: B
Rationale: Approximate nearest neighbor indexing speeds up retrieval.
48.
A developer wants scalable AI pipelines. What is BEST?
A. Manual
B. Step Functions
C. EC2
D. S3
Answer: B
Rationale: Step Functions orchestrate workflows.
49.
A developer wants monitoring. What is BEST?
A. CloudWatch
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: CloudWatch monitors metrics.
50.
A developer wants log analysis. What is BEST?
A. CloudWatch Logs Insights
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: Logs Insights queries logs.
51.
A developer wants CI/CD for AI apps. What is BEST?
A. CodePipeline
B. EC2
C. S3
D. RDS
Answer: A
Rationale: CodePipeline automates deployments.
52.
A developer wants model monitoring. What is BEST?
A. CloudWatch
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: CloudWatch tracks performance.
53.
A developer wants event-driven AI workflows. What is BEST?
A. EventBridge
B. EC2
C. RDS
D. S3
Answer: A
Rationale: EventBridge triggers workflows.
54.
A developer wants scalable storage. What is BEST?
A. S3
B. EC2
C. RDS
D. DynamoDB
Answer: A
Rationale: S3 scales automatically.
55.
A developer wants authentication. What is BEST?
A. IAM
B. Cognito
C. S3
D. EC2
Answer: B
Rationale: Cognito manages users.
56.
A developer wants encryption. What is BEST?
A. IAM
B. KMS
C. CloudWatch
D. Lambda
Answer: B
Rationale: KMS manages encryption.
57.
A developer wants high availability. What is BEST?
A. Single AZ
B. Multi-AZ
C. EC2
D. S3
Answer: B
Rationale: Multi-AZ ensures redundancy.
58.
A developer wants cost optimization. What is BEST?
A. Large models only
B. Use smaller models when possible
C. EC2
D. S3
Answer: B
Rationale: Smaller models reduce cost.
59.
A developer wants scalable APIs. What is BEST?
A. API Gateway + Lambda
B. EC2
C. S3
D. RDS
Answer: A
Rationale: Serverless APIs scale automatically.
60.
A developer wants production-ready AI app. What is BEST?
A. Single service
B. Bedrock + API Gateway + Lambda + monitoring
C. EC2
D. S3
Answer: B
Rationale: A full serverless architecture ensures scalability, security, and observability for production GenAI applications.
61.
A RAG system retrieves outdated documents. What is BEST?
A. Increase tokens
B. Implement document versioning and freshness filtering
C. Use EC2
D. Use S3
Answer: B
Rationale: Retrieval pipelines must include metadata like timestamps and versioning. Filtering based on freshness ensures only relevant and up-to-date content is used for generation.
62.
A developer wants to reduce embedding storage costs. What is BEST?
A. Increase embeddings
B. Use dimensionality reduction or compression
C. EC2
D. S3
Answer: B
Rationale: Reducing embedding dimensionality or compressing vectors lowers storage costs while maintaining acceptable retrieval accuracy.
63.
A chatbot gives inconsistent answers to the same query. What is BEST?
A. Increase temperature
B. Reduce temperature and standardize prompts
C. EC2
D. S3
Answer: B
Rationale: Lower temperature reduces randomness, and consistent prompt templates ensure stable outputs.
64.
A developer wants to prevent sensitive data leakage. What is BEST?
A. Ignore
B. Data masking and access controls
C. EC2
D. S3
Answer: B
Rationale: Masking and strict IAM policies prevent exposure of sensitive information in prompts or outputs.
65.
A RAG system has slow retrieval latency. What is BEST?
A. Linear search
B. Use ANN indexing in vector DB
C. EC2
D. S3
Answer: B
Rationale: Approximate nearest neighbor indexing significantly improves retrieval speed.
66.
A developer wants to evaluate hallucination rates. What is BEST?
A. Ignore
B. Ground truth comparison and evaluation datasets
C. EC2
D. S3
Answer: B
Rationale: Comparing outputs to known correct answers helps quantify hallucinations.
67.
A developer wants multi-modal AI (text + image). What is BEST?
A. S3
B. Bedrock multi-modal model
C. EC2
D. RDS
Answer: B
Rationale: Bedrock supports multi-modal foundation models for text and images.
68.
A developer wants scalable prompt experimentation. What is BEST?
A. Hardcode
B. A/B testing framework
C. EC2
D. S3
Answer: B
Rationale: A/B testing allows comparison of prompt variations to optimize outputs.
69.
A developer wants to handle long documents in RAG. What is BEST?
A. Single chunk
B. Chunking with overlap
C. EC2
D. S3
Answer: B
Rationale: Chunking with overlap preserves context across segments and improves retrieval accuracy.
70.
A developer wants to minimize token usage. What is BEST?
A. Increase context
B. Prompt compression and summarization
C. EC2
D. S3
Answer: B
Rationale: Reducing prompt size lowers cost and improves efficiency.
71.
A developer wants real-time monitoring of LLM errors. What is BEST?
A. CloudTrail
B. CloudWatch metrics and logs
C. Config
D. Lambda
Answer: B
Rationale: CloudWatch enables real-time monitoring and alerting for AI systems.
72.
A developer wants to detect prompt injection attempts. What is BEST?
A. Ignore
B. Input validation + anomaly detection
C. EC2
D. S3
Answer: B
Rationale: Filtering and anomaly detection help prevent malicious prompt manipulation.
73.
A developer wants to scale embedding generation. What is BEST?
A. Manual
B. Parallel batch processing
C. EC2
D. S3
Answer: B
Rationale: Parallel processing increases throughput for embedding generation.
74.
A developer wants secure model access. What is BEST?
A. Public access
B. IAM policies + private endpoints
C. EC2
D. S3
Answer: B
Rationale: IAM and private networking secure model endpoints.
75.
A developer wants to improve answer accuracy. What is BEST?
A. Increase temperature
B. Use RAG with curated data
C. EC2
D. S3
Answer: B
Rationale: Curated data improves reliability.
76.
A developer wants to store conversation history. What is BEST?
A. Ignore
B. DynamoDB or database storage
C. EC2
D. S3
Answer: B
Rationale: Persistent storage enables context-aware conversations.
77.
A developer wants streaming responses. What is BEST?
A. Batch
B. Streaming API
C. EC2
D. S3
Answer: B
Rationale: Streaming improves UX.
78.
A developer wants automated pipelines. What is BEST?
A. Manual
B. Step Functions
C. EC2
D. S3
Answer: B
Rationale: Step Functions orchestrate workflows.
79.
A developer wants model evaluation automation. What is BEST?
A. Manual
B. Automated evaluation pipelines
C. EC2
D. S3
Answer: B
Rationale: Automation ensures consistent evaluation.
80.
A developer wants vector DB optimization. What is BEST?
A. No index
B. Index tuning
C. EC2
D. S3
Answer: B
Rationale: Index tuning improves performance.
81.
A developer wants data privacy compliance. What is BEST?
A. Ignore
B. Encryption + access control
C. EC2
D. S3
Answer: B
Rationale: Protects sensitive data.
82.
A developer wants API scaling. What is BEST?
A. EC2
B. API Gateway + Lambda
C. S3
D. RDS
Answer: B
Rationale: Serverless APIs scale automatically.
83.
A developer wants AI workflow automation. What is BEST?
A. EventBridge
B. EC2
C. RDS
D. S3
Answer: A
Rationale: EventBridge triggers workflows.
84.
A developer wants cost monitoring. What is BEST?
A. CloudWatch
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: CloudWatch tracks usage.
85.
A developer wants logging. What is BEST?
A. CloudWatch Logs
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: Logs enable debugging.
86.
A developer wants CI/CD. What is BEST?
A. CodePipeline
B. EC2
C. S3
D. RDS
Answer: A
Rationale: CodePipeline automates deployments.
87.
A developer wants secure secrets. What is BEST?
A. Hardcode
B. Secrets Manager
C. S3
D. EC2
Answer: B
Rationale: Secrets Manager stores securely.
88.
A developer wants encryption. What is BEST?
A. IAM
B. KMS
C. CloudWatch
D. Lambda
Answer: B
Rationale: KMS manages encryption.
89.
A developer wants high availability. What is BEST?
A. Single AZ
B. Multi-AZ
C. EC2
D. S3
Answer: B
Rationale: Multi-AZ ensures redundancy.
90.
A developer wants production-ready GenAI system. What is BEST?
A. Single service
B. Bedrock + RAG + API Gateway + monitoring
C. EC2
D. S3
Answer: B
Rationale: A full architecture ensures scalability, accuracy, and observability.
91.
A RAG system retrieves correct documents but answers are still incorrect. What is BEST?
A. Increase retrieval
B. Improve prompt grounding instructions
C. Use EC2
D. Use S3
Answer: B
Rationale: Even with correct retrieval, poor prompt instructions can cause the LLM to ignore context. Strong grounding prompts ensure the model uses retrieved data accurately.
92.
A developer wants to evaluate prompt performance at scale. What is BEST?
A. Manual testing
B. Automated evaluation pipeline with datasets
C. EC2
D. S3
Answer: B
Rationale: Automated pipelines enable consistent, repeatable evaluation across large datasets, improving prompt optimization.
93.
A developer wants to reduce token costs across millions of requests. What is BEST?
A. Increase context
B. Prompt compression + response truncation
C. EC2
D. S3
Answer: B
Rationale: Reducing token size directly lowers cost, especially at scale.
94.
A developer wants hybrid search (semantic + keyword). What is BEST?
A. Only embeddings
B. Combine vector search with keyword search
C. EC2
D. S3
Answer: B
Rationale: Hybrid search improves retrieval accuracy by combining semantic similarity with keyword matching.
95.
A developer wants to reduce latency in RAG pipelines. What is BEST?
A. Increase tokens
B. Cache embeddings and retrieval results
C. EC2
D. S3
Answer: B
Rationale: Caching reduces repeated computations and improves response times.
96.
A developer wants to detect model drift. What is BEST?
A. Ignore
B. Continuous evaluation with baseline comparison
C. EC2
D. S3
Answer: B
Rationale: Comparing outputs over time detects drift and performance degradation.
97.
A developer wants to orchestrate multi-step AI workflows. What is BEST?
A. Manual
B. Step Functions
C. EC2
D. S3
Answer: B
Rationale: Step Functions coordinate complex workflows across services.
98.
A developer wants to prevent sensitive data in prompts. What is BEST?
A. Ignore
B. Input filtering and redaction
C. EC2
D. S3
Answer: B
Rationale: Filtering prevents leakage of sensitive data.
99.
A developer wants scalable vector search. What is BEST?
A. Linear search
B. Managed vector database
C. EC2
D. S3
Answer: B
Rationale: Managed vector DB scales efficiently.
100.
A developer wants to test multiple models. What is BEST?
A. Single model
B. Model comparison framework
C. EC2
D. S3
Answer: B
Rationale: Comparing models helps select best performer.
101.
A developer wants multi-agent AI systems. What is BEST?
A. Single agent
B. Orchestrated agents with workflows
C. EC2
D. S3
Answer: B
Rationale: Multi-agent systems handle complex tasks collaboratively.
102.
A developer wants to monitor hallucinations. What is BEST?
A. Ignore
B. Evaluation datasets + scoring
C. EC2
D. S3
Answer: B
Rationale: Metrics track hallucination rates.
103.
A developer wants real-time moderation. What is BEST?
A. Ignore
B. Content moderation API
C. EC2
D. S3
Answer: B
Rationale: Moderation APIs filter harmful content.
104.
A developer wants retrieval ranking improvement. What is BEST?
A. Random
B. Re-ranking models
C. EC2
D. S3
Answer: B
Rationale: Re-ranking improves relevance.
105.
A developer wants cost governance. What is BEST?
A. Ignore
B. Usage monitoring + limits
C. EC2
D. S3
Answer: B
Rationale: Monitoring prevents overspending.
106.
A developer wants prompt security. What is BEST?
A. Ignore
B. Guardrails and validation
C. EC2
D. S3
Answer: B
Rationale: Guardrails prevent misuse.
107.
A developer wants scalable inference. What is BEST?
A. EC2
B. Serverless inference APIs
C. S3
D. RDS
Answer: B
Rationale: Serverless scales automatically.
108.
A developer wants batch embeddings. What is BEST?
A. Manual
B. Batch processing pipeline
C. EC2
D. S3
Answer: B
Rationale: Batch improves efficiency.
109.
A developer wants logging. What is BEST?
A. CloudWatch Logs
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: Logs help debugging.
110.
A developer wants monitoring alerts. What is BEST?
A. CloudWatch alarms
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: Alerts notify issues.
111.
A developer wants CI/CD. What is BEST?
A. CodePipeline
B. EC2
C. S3
D. RDS
Answer: A
Rationale: CodePipeline automates deployments.
112.
A developer wants secure secrets. What is BEST?
A. Hardcode
B. Secrets Manager
C. S3
D. EC2
Answer: B
Rationale: Secure storage.
113.
A developer wants encryption. What is BEST?
A. IAM
B. KMS
C. CloudWatch
D. Lambda
Answer: B
Rationale: Encryption.
114.
A developer wants scalable APIs. What is BEST?
A. API Gateway + Lambda
B. EC2
C. S3
D. RDS
Answer: A
Rationale: Serverless APIs scale.
115.
A developer wants workflow automation. What is BEST?
A. EventBridge
B. EC2
C. RDS
D. S3
Answer: A
Rationale: Event-driven automation.
116.
A developer wants data storage. What is BEST?
A. S3
B. EC2
C. RDS
D. DynamoDB
Answer: A
Rationale: Scalable storage.
117.
A developer wants authentication. What is BEST?
A. IAM
B. Cognito
C. S3
D. EC2
Answer: B
Rationale: User auth.
118.
A developer wants high availability. What is BEST?
A. Single AZ
B. Multi-AZ
C. EC2
D. S3
Answer: B
Rationale: Redundancy.
119.
A developer wants cost optimization. What is BEST?
A. Large models
B. Smaller models + caching
C. EC2
D. S3
Answer: B
Rationale: Cost-efficient strategy.
120.
A developer wants production GenAI architecture. What is BEST?
A. Single service
B. Bedrock + RAG + vector DB + API Gateway + monitoring
C. EC2
D. S3
Answer: B
Rationale: Full architecture ensures scalability, reliability, and accuracy.
121.
A RAG system retrieves too many irrelevant chunks. What is BEST?
A. Increase chunk size
B. Improve chunking and filtering strategy
C. EC2
D. S3
Answer: B
Rationale: Poor chunking leads to noisy retrieval. Optimizing chunk size, overlap, and metadata filtering improves precision and relevance in RAG systems.
122.
A developer wants to prioritize most relevant results. What is BEST?
A. Random order
B. Re-ranking model after retrieval
C. EC2
D. S3
Answer: B
Rationale: Re-ranking models reorder retrieved results based on relevance, improving final output quality.
123.
A developer wants to reduce hallucinations in domain-specific apps. What is BEST?
A. Increase temperature
B. Fine-tune model + RAG
C. EC2
D. S3
Answer: B
Rationale: Combining fine-tuning with RAG improves accuracy and reduces hallucinations.
124.
A developer wants secure prompt handling. What is BEST?
A. Ignore
B. Input validation and sanitization
C. EC2
D. S3
Answer: B
Rationale: Sanitization prevents malicious input.
125.
A developer wants to scale RAG pipelines. What is BEST?
A. Manual
B. Distributed architecture with caching
C. EC2
D. S3
Answer: B
Rationale: Distributed systems handle scale efficiently.
126.
A developer wants evaluation automation. What is BEST?
A. Manual
B. Automated evaluation pipelines
C. EC2
D. S3
Answer: B
Rationale: Automation ensures consistency.
127.
A developer wants to detect bias in outputs. What is BEST?
A. Ignore
B. Evaluation datasets + fairness metrics
C. EC2
D. S3
Answer: B
Rationale: Metrics detect bias.
128.
A developer wants to reduce latency globally. What is BEST?
A. Single region
B. Edge caching and regional endpoints
C. EC2
D. S3
Answer: B
Rationale: Edge caching improves performance.
129.
A developer wants scalable vector storage. What is BEST?
A. RDS
B. Managed vector DB
C. S3
D. EC2
Answer: B
Rationale: Vector DB scales.
130.
A developer wants multi-agent workflows. What is BEST?
A. Single agent
B. Orchestrated multi-agent system
C. EC2
D. S3
Answer: B
Rationale: Multi-agent systems solve complex tasks.
131.
A developer wants prompt version control. What is BEST?
A. Hardcode
B. Versioning system
C. EC2
D. S3
Answer: B
Rationale: Versioning enables tracking.
132.
A developer wants model monitoring. What is BEST?
A. CloudWatch
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: Monitoring tracks performance.
133.
A developer wants logging. What is BEST?
A. CloudWatch Logs
B. CloudTrail
C. Config
D. Lambda
Answer: A
Rationale: Logs enable debugging.
134.
A developer wants CI/CD. What is BEST?
A. CodePipeline
B. EC2
C. S3
D. RDS
Answer: A
Rationale: CI/CD automates deployments.
135.
A developer wants secrets security. What is BEST?
A. Hardcode
B. Secrets Manager
C. S3
D. EC2
Answer: B
Rationale: Secure storage.
136.
A developer wants encryption. What is BEST?
A. IAM
B. KMS
C. CloudWatch
D. Lambda
Answer: B
Rationale: Encryption keys.
137.
A developer wants scalable APIs. What is BEST?
A. API Gateway + Lambda
B. EC2
C. S3
D. RDS
Answer: A
Rationale: Serverless APIs scale.
138.
A developer wants workflow automation. What is BEST?
A. EventBridge
B. EC2
C. RDS
D. S3
Answer: A
Rationale: Event-driven automation.
139.
A developer wants storage. What is BEST?
A. S3
B. EC2
C. RDS
D. DynamoDB
Answer: A
Rationale: Scalable storage.
140.
A developer wants authentication. What is BEST?
A. IAM
B. Cognito
C. S3
D. EC2
Answer: B
Rationale: User auth.
141.
A developer wants high availability. What is BEST?
A. Single AZ
B. Multi-AZ
C. EC2
D. S3
Answer: B
Rationale: Redundancy.
142.
A developer wants cost optimization. What is BEST?
A. Large models
B. Smaller models + caching
C. EC2
D. S3
Answer: B
Rationale: Cost-efficient.
143.
A developer wants evaluation metrics. What is BEST?
A. Ignore
B. Precision/recall + human review
C. EC2
D. S3
Answer: B
Rationale: Metrics ensure quality.
144.
A developer wants retrieval improvement. What is BEST?
A. Random
B. Hybrid search
C. EC2
D. S3
Answer: B
Rationale: Hybrid improves accuracy.
145.
A developer wants prompt safety. What is BEST?
A. Ignore
B. Guardrails
C. EC2
D. S3
Answer: B
Rationale: Guardrails ensure safety.
146.
A developer wants inference scaling. What is BEST?
A. EC2
B. Serverless inference
C. S3
D. RDS
Answer: B
Rationale: Serverless scales.
147.
A developer wants batch processing. What is BEST?
A. Manual
B. Batch pipelines
C. EC2
D. S3
Answer: B
Rationale: Efficient processing.
148.
A developer wants moderation. What is BEST?
A. Ignore
B. Moderation models
C. EC2
D. S3
Answer: B
Rationale: Filters harmful content.
149.
A developer wants anomaly detection. What is BEST?
A. CloudTrail
B. CloudWatch anomaly detection
C. Config
D. Lambda
Answer: B
Rationale: Detects anomalies.
150.
A developer wants production-ready GenAI system. What is BEST?
A. Single service
B. Bedrock + RAG + vector DB + API Gateway + monitoring + guardrails
C. EC2
D. S3
Answer: B
Rationale: A complete architecture ensures scalability, safety, observability, and high-quality outputs in real-world GenAI applications.