Study guide
7.1 Managed AI Services vs Foundation Model Apps vs Custom ML
AWS AI Practitioner Study Guide Exam
Track your progress as you read
Finished sections get a check mark.
1 Chapter 1: AIF-C01 Orientation and Official Source Control
1.1 Credential Scope, Code, and Search Language 1.2 Exam Format, Cost, Score, and Delivery 1.3 Official Exam Guide, Skill Builder, and Practice Workflow 1.4 Target Candidate Boundaries and Out-of-Scope Tasks 1.5 Domain Weights and Study Prioritization 1.6 Exam Policy, Retake, Results, and Recertification
2 Chapter 2: AI/ML Foundations and Use-Case Fit
2.1 AI, ML, Deep Learning, and Core Terminology 2.2 Supervised, Unsupervised, and Reinforcement Learning 2.3 Data Types, Labels, Structure, and Quality 2.4 Inference Patterns: Batch, Real-Time, and Embedded AI 2.5 AI Use-Case Fit and No-AI Decisions 2.6 Classification, Regression, Clustering, Forecasting, and Recommendation 2.7 AI/ML Foundations Case Lab
3 Chapter 3: ML Lifecycle, Metrics, and Practitioner MLOps
3.1 Data Collection, EDA, Preprocessing, and Feature Concepts 3.2 Training, Evaluation, Deployment, and Monitoring Lifecycle 3.3 Model Sources: Managed APIs, Open Source, and Custom Models 3.4 Evaluation Metrics and Business Metrics 3.5 Practitioner MLOps: Repeatability, Monitoring, and Retraining 3.6 SageMaker Lifecycle Service Map 3.7 ML Lifecycle Case Lab
4 Chapter 4: Generative AI Foundations and Inference Concepts
4.1 Foundation Models, LLMs, Transformers, and Modalities 4.2 Tokens, Context Windows, Embeddings, and Vector Search 4.3 Inference Parameters: Temperature, Top-p, and Output Controls 4.4 Hallucination, Grounding, and Context Quality 4.5 GenAI Use-Case Fit and Risk Triage 4.6 Prompt Injection, Data Leakage, and User Input Risk 4.7 Generative AI Foundations Case Lab
5 Chapter 5: Prompting, Model Selection, Customization, and Evaluation
5.1 Prompt Engineering Patterns and Business Quality 5.2 Zero-Shot, Few-Shot, Templates, and Instruction Design 5.3 Model Selection: Capability, Latency, Cost, and Risk 5.4 RAG vs Fine-Tuning vs Prompting vs Custom Models 5.5 Model Evaluation, Human Review, and Red-Team Feedback 5.6 Cost, Performance, and Throughput Decision-Making 5.7 Prompting and Model Selection Case Lab
6 Chapter 6: Amazon Bedrock, RAG, Agents, and Guardrails
6.1 Bedrock Core Concepts, Model Access, and Managed FM Choice 6.2 Knowledge Bases, RAG, Data Sources, and Grounding 6.3 AI Agents, Tools, Orchestration, and Business Workflows 6.4 Guardrails, Content Filters, Denied Topics, and Sensitive Data 6.5 Bedrock Model Evaluation, Monitoring, and Human Feedback 6.6 Bedrock Cost, Latency, Throughput, and Operational Fit 6.7 Bedrock, RAG, and Agents Case Lab
7 Chapter 7: AWS Managed AI/ML Services and SageMaker Map
7.1 Managed AI Services vs Foundation Model Apps vs Custom ML 7.2 Text, Language, Search, and Document AI Services 7.3 Vision, Speech, Contact Center, and Personalization Services 7.4 Amazon Quick, Amazon Q Developer, Kiro, and Practitioner Fit 7.5 SageMaker Canvas, Studio, Clarify, Autopilot, and Data Wrangler 7.6 Data Foundation Services: S3, Glue, OpenSearch, and Quick Sight 7.7 AWS AI Service Selection Case Lab
8 Chapter 8: Responsible AI, Human Review, and Safety
8.1 Fairness, Bias, Transparency, and Explainability 8.2 Privacy, Safety, Human Review, and Accountability 8.3 Guardrails, Clarify, Human Review, and Content Safety Controls 8.4 Responsible AI Risk Registers and Governance Workflows 8.5 Monitoring, Feedback, Drift, and Incident Response 8.6 Responsible AI Case Lab
9 Chapter 9: Security, Compliance, Governance, and Cost Controls
9.1 Shared Responsibility, IAM, and Least Privilege for AI 9.2 Encryption, Secrets, Networking, and Data Privacy 9.3 Prompt Injection, Data Exfiltration, and GenAI Threat Modeling 9.4 Logging, Monitoring, CloudTrail, CloudWatch, and Config 9.5 Compliance: Artifact, Config, Macie, and Policy Evidence 9.6 AI Cost Controls, Pricing, Throughput, and Budget Governance 9.7 Security and Governance Case Lab
10 Chapter 10: Integrated AWS AI Business Scenario Labs
10.1 Customer Support GenAI Assistant Lab 10.2 Document Intelligence and Compliance Review Lab 10.3 Personalization, Forecasting, and Fraud Detection Lab 10.4 Enterprise Search, RAG, and Knowledge Management Lab 10.5 Responsible AI and Security Review Board Lab 10.6 Cost, Performance, and Operations Review Lab 10.7 Full AIF-C01 Business Simulation
11 Chapter 11: Final Review, Exam Readiness, and Recertification
11.1 Final 30-Day AIF-C01 Study Plan 11.2 Official Practice Resources and Weak-Domain Remediation 11.3 90-Minute Exam Timing, Flagging, and Guessing Workflow 11.4 Test-Day Checklist: Online or Test Center 11.5 Post-Exam Results, Retake, and Recertification Plan 11.6 AWS AI Practitioner Final Mixed Review
7.1 Managed AI Services vs Foundation Model Apps vs Custom ML
Key Takeaways Explain This Topic AI
Managed AI services are usually the fastest fit when the task is a common capability such as transcription, translation, document extraction, vision analysis, search, or personalization.
Foundation model applications fit open-ended generation, chat, summarization, RAG, assistants, and agentic workflows where probabilistic output is acceptable with controls.
Custom ML with Amazon SageMaker AI fits unique prediction problems, proprietary model training, deeper feature work, and teams that need lifecycle control.
Practitioners should select the smallest ownership boundary that meets the business goal, data requirements, risk tolerance, and operating model.
Not every automation needs AI; deterministic rules, workflow tools, or analytics can be better when outputs must be exact and explainable.
Last updated: October 2026
Choosing the right AI ownership layer
A practitioner does not need to build neural networks to make good AWS AI decisions. The practical skill is to map the business problem to the lightest AWS service layer that can solve it responsibly. AWS offers managed AI services for common tasks, foundation model application services for generative AI workflows, and SageMaker AI for custom ML work. Those layers have different costs, risks, data needs, and operating responsibilities.
Managed AI services are prebuilt capabilities exposed through AWS APIs and consoles. Examples include Amazon Textract for document text and structure extraction, Amazon Transcribe for speech-to-text, Amazon Translate for language translation, Amazon Comprehend for natural language insights, Amazon Rekognition for image and video analysis, and Amazon Personalize for recommendations. The team supplies data and workflow context, but does not train the underlying model.
Foundation model applications are different. A team uses a model or assistant to generate, summarize, classify, reason over retrieved context, or help users take action. Amazon Bedrock is the managed foundation model layer for builders. Amazon Quick and Amazon Q Developer are managed assistant experiences. These options can be powerful, but they are probabilistic and need grounding, permission design, evaluation, logging decisions, and human review for risky workflows.
Custom ML is the deeper build path. Amazon SageMaker AI supports teams that need to prepare data, build or tune models, run experiments, train, host, monitor, and manage the ML lifecycle. AIF-C01 candidates are not expected to implement those pipelines, tune hyperparameters, or code algorithms. They should still know when a business requirement crosses from consuming a managed capability into owning a custom ML lifecycle.
Service path Best fit Practitioner boundary question
Managed AI service Common task with a prebuilt AWS API Is the task standard enough that a managed service can do it without custom training?
Foundation model app Chat, generation, summarization, RAG, assistant, or agent workflow Can the business tolerate probabilistic outputs with grounding and controls?
Custom ML with SageMaker AI Unique prediction, proprietary model, lifecycle control, specialized data science work Is the team ready to own data prep, training, evaluation, deployment, and monitoring decisions?
No AI or simple automation Exact rules, deterministic workflow, simple reporting, or low value use case Would a rule, SQL query, dashboard, or workflow approval solve the problem more safely?
The first service-selection question is the business outcome. Do users need a transcript, a translated document, a search answer, a recommendation, a prediction, a generated response, or a workflow action? Each output type points to a different AWS family. A transcript points to Transcribe. A generated support reply might point to Bedrock or Amazon Quick. A custom churn prediction using company labels might point to SageMaker AI or SageMaker Canvas.
The second question is whether the task is standard. Extracting forms from invoices is a common document AI need, so Textract is a strong starting point. Finding named entities or sentiment in support tickets is a common NLP need, so Comprehend might fit. Building a model to predict failures from a proprietary sensor stream is less standard because labels, equipment behavior, and costs are company-specific. That type of requirement often needs a custom ML discussion.
The third question is data and governance. Managed AI services still process customer data, so the team must evaluate IAM, encryption, retention, logs, privacy, Region choice, and whether human review is needed. Foundation model apps add prompt injection, hallucination, source quality, and output safety concerns. Custom ML adds training data lineage, bias evaluation, model drift, retraining, and deployment ownership.
Service selection should also avoid overbuilding. A team that wants to search approved policies may not need to train a model. A team that wants employees to ask questions over internal documents may start with Amazon Quick or a Bedrock RAG pattern. A team that wants a regulated score used for credit, hiring, or safety decisions needs a higher bar for fairness, explainability, human review, and monitoring before any AI path is approved.
Use this approval checklist before recommending a path:
Define the exact user decision or task the AI output will support.
Identify the data type: text, image, audio, document, tabular, time-series, clickstream, or mixed data.
Decide whether the task is a common AWS managed AI capability or a unique company-specific prediction problem.
Confirm whether probabilistic output is acceptable, and where human review is required.
Ask who will own data quality, permissions, model evaluation, monitoring, cost review, and user feedback.
Prefer AWS Skill Builder labs and official practice workflows to explore service boundaries in a sandbox before production approval.
A good practitioner recommendation is often conservative. Start with a managed service when it directly matches the task. Use a foundation model application when language generation, reasoning over context, or assistant behavior is the actual requirement. Move to SageMaker AI only when the business truly needs custom model ownership and has a team capable of operating the lifecycle.