AWS AIF-C01 Cheat Sheet: GenAI, RAG, Agents, and Bedrock

Keep this cheat sheet open while drilling questions. AIF‑C01 rewards clean definitions, best-fit service selection, and risk-aware design (hallucinations, privacy, prompt injection, responsible use).

Quick facts (AIF-C01)

I verified these current AWS exam facts on May 24, 2026.

Domain weights and review priority

The highest-weight domain is Applications of Foundation Models. Spend final review time on service fit, RAG, prompt patterns, and model-risk decisions, not on deep ML implementation details that AWS says are out of scope for the target candidate.

Practitioner scope boundary

AIF-C01 is a foundational use-and-decide exam, not an ML engineering exam. When an answer asks you to build, tune, or operate a full ML platform, check whether the stem really requires that depth.

Keep the answer at this altitude:

1. Business fit: What outcome does the organization need?
2. AI pattern fit: Is this classification, prediction, GenAI, RAG, agentic tool use, extraction, or evaluation?
3. AWS service fit: Which managed AWS service or feature matches the use case with the least unnecessary complexity?
4. Risk fit: What could go wrong: hallucination, bias, prompt injection, sensitive-data exposure, cost, latency, or weak governance?
5. Control fit: Which control closes the risk: IAM, encryption, redaction, guardrails, citations, human review, monitoring, or audit evidence?

Avoid answers that jump straight to model training, feature engineering, hyperparameter tuning, or custom pipeline infrastructure unless the question explicitly asks for those advanced implementation details.

Official task compression

Current objective cues

Treat these as current final-review cues from the AWS exam guide:

Final answer stack

When two answers both sound plausible, prefer the one that follows this stack:

1. Define the use case before the service. Do not choose Bedrock, SageMaker AI, Comprehend, Textract, or Rekognition until the task is clear.
2. Choose managed capability before custom build. AIF-C01 usually rewards the right managed service, not custom ML infrastructure.
3. Ground facts before tuning style. If the problem is stale or private knowledge, fix retrieval and citations before changing model size.
4. Protect data before improving output. Sensitive prompts, retrieved context, logs, and generated output all need data-boundary thinking.
5. Evaluate before trusting. Use representative tests, feedback, monitoring, and review for quality, safety, and responsible use.

How AIF-C01 questions work (fast strategy)

• Treat every question as scored. Unanswered questions are wrong, and there is no penalty for guessing.
• For ordering and matching, all placements/pairs must be correct to get credit.
• If the prompt says least operational effort, prefer managed services (and native integrations).
• If the question is about “improving factual accuracy,” the best answer is often grounding (RAG + citations) rather than “make the model bigger.”
• If the question is about “sensitive data,” the best answer usually includes least privilege, encryption, and minimizing what you send to the model.
• If the scenario includes “untrusted user input,” think prompt injection defenses and safe tool use (allowlists, scoped permissions).
• Read the last sentence first to capture the constraint (cost, latency, safety, compliance).

Question-type traps

For ordering and matching, all placements or pairs must be correct to receive credit. Do not rush these as if they were ordinary single-answer items.

Final 20-minute recall (exam day)

Cue -> best answer (pattern map)

Practitioner decision chooser

Must-memorize formulas and levers

Last-minute traps

• Picking fine-tuning when RAG + prompt engineering already solves grounding.
• Sending raw PII/secrets to model prompts or logs.
• Ignoring prompt injection in agent/tool scenarios.
• Treating model confidence as truth without retrieval evidence.

Target candidate and scope boundaries

• Target depth is practitioner-level: up to ~6 months of AI/ML exposure on AWS.
• You are expected to use AI/ML services, not build custom AI/ML systems from scratch.
• Usually out of scope: coding models, feature engineering, hyperparameter tuning, building ML pipelines/infrastructure, deep math/statistics, and designing governance frameworks from zero.

0) Core mental model: a GenAI app (with RAG)

    flowchart LR
	  U[User] --> A[App / API]
	  A -->|Prompt + context| FM[Foundation Model]
	  A -->|Embed query| E[Embeddings]
	  E --> VS[(Vector Store)]
	  VS -->|Top-k chunks| A
	  A -->|Policy filters| G[Guardrails / Moderation]
	  A -->|Logs/metrics| O[Observability]
	  A -->|AuthN/AuthZ| IAM[IAM / Identity]

RAG in one sentence: retrieve relevant private content, then ask the model to answer using only that content (ideally with citations).

Pattern chooser

• Prompting when the issue is wording, format, tone, or instructions.
• RAG when the issue is stale, private, or enterprise knowledge.
• Agent when the issue is multi-step tool use with permissions and audit.
• Guardrails when the issue is unsafe output, policy, or data exposure.
• Fine-tuning only when the behavior is stable enough to learn from examples.

Prompt, RAG, agent, or fine-tuning?

Scenario eliminations

The practitioner exam rewards “right managed capability for the business use case.” If an answer asks you to build, tune, or operate a full ML platform without a clear need, it is often too advanced for AIF-C01.

AIF-C01 answer sequence

Use this when the stem mixes model choice, retrieval, safety, and operations.

    flowchart TD
	  S["Scenario"] --> W["Workload and goal"]
	  W --> R["Risk: stale facts, unsafe output, exposure, latency, or cost"]
	  R --> P["Pick prompt, RAG, agent, guardrail, or workflow"]
	  P --> C["Apply IAM, KMS, logging, review, retry, or rollback"]
	  C --> E["Verify with citations, metrics, tests, or audit trail"]

1) AI/ML fundamentals (Domain 1)

Core terminology (must know)

Metrics (common, conceptual)

ML lifecycle (high level)

    flowchart LR
	  P[Define problem + metric] --> D[Collect/prepare data]
	  D --> T[Train + tune]
	  T --> E[Evaluate]
	  E --> DEP[Deploy]
	  DEP --> M[Monitor + feedback]
	  M --> D

Common best answer patterns:

• If you can’t define a metric or get data, ML is usually the wrong first move.
• Production ML needs monitoring (quality/latency/cost) and retraining plans.

2) Generative AI fundamentals (Domain 2)

Key GenAI terms (must know)

Prompting vs RAG vs fine-tuning (decision table)

Context and token-cost levers

GenAI limitations to recognize

• Factuality isn’t guaranteed → use grounding/citations and “unknown” responses.
• Context is limited → don’t paste entire corpora; retrieve and summarize.
• Outputs can be unsafe/biased → add guardrails, evaluation, and human review paths.
• Costs scale with tokens → control prompt size, choose smaller models when acceptable, cache repeated work.

3) AWS service map (what to pick when)

Foundation models and ML platforms

Bedrock versus SageMaker AI

For AIF-C01, Bedrock usually means use foundation models. SageMaker AI usually means build or manage ML model lifecycle. Do not pick SageMaker AI just because it sounds more powerful.

Bedrock features to recognize quickly

Agentic AI risk map

Pre-built AI services (use-case driven)

Common building blocks for GenAI apps (glue)

4) RAG: design notes that show up in exam scenarios (Domain 3)

RAG architecture (end-to-end)

    flowchart TB
	  subgraph Ingestion
	    S3[(Docs in S3)] --> C[Chunk + clean]
	    C --> EMB1[Create embeddings]
	    EMB1 --> VS[(Vector store)]
	  end
	
	  subgraph Answering
	    Q[User question] --> EMB2[Embed query]
	    EMB2 --> VS
	    VS --> K[Top-k chunks]
	    K --> P[Prompt template: instructions + context]
	    P --> FM[Foundation model]
	    FM --> A[Answer + citations]
	  end

High-yield design choices

• Chunking: smaller chunks improve precision; larger chunks improve context. The exam often wants “tune chunking for relevance.”
• Citations: if the requirement says “trust” or “audit,” add citations/source links.
• Freshness: if content changes often, prefer RAG over fine-tuning.
• Privacy: don’t send more data than needed; redact PII; restrict who can retrieve what (multi-tenant boundaries).

5) Prompt engineering patterns (Domain 3)

Techniques you should recognize

Prompt template (practical)

 1Goal: Answer the user question using ONLY the provided context.
 2Context:
 3<<<
 4{retrieved_chunks}
 5>>>
 6Rules:
 7- If the answer is not in the context, say "Insufficient context".
 8- Provide 2-3 bullet citations (source titles/ids).
 9Output format (JSON):
10{"answer":"...", "citations":[{"source":"...","quote":"..."}]}
11User question: {question}

Anti-prompt-injection rule of thumb

Treat user-provided text as data, not instructions. If the model is allowed to call tools/actions, use allowlists and scoped permissions.

6) Evaluation and monitoring (Domain 3)

What to evaluate

Common “best answers”:

• Use a representative test set (not just a few demos).
• Do A/B testing when changing prompts/models.
• Monitor production for quality regressions and abuse.

Evaluation method chooser

7) Responsible AI (Domain 4)

Responsible AI checklist (high signal)

• Define intended use + out-of-scope use (avoid “silent expansion”).
• Add human oversight for high-impact decisions.
• Evaluate for bias and document limitations.
• Implement safety policies (harmful content, privacy leakage).
• Be transparent with users (what it is, what it isn’t, how to verify).

Common risks and mitigations

8) Security, compliance, and governance (Domain 5)

Security “gotchas” the exam expects you to notice

• Over-permissive IAM roles (“*” actions/resources)
• Secrets embedded in prompts, logs, or code
• Sending unnecessary sensitive data to the model
• No audit trail for access and changes
• Tool use without constraints (model can “do anything”)

AWS controls to name in answers (by theme)

Security answer bundle

Service-name exam tip

• The exam uses AWS short service names in questions (for example, Amazon SNS instead of a longer formal name).
• Use the exam Help pane for service-name expansion when available.

Next steps

• Use Resources to work directly from the official exam guide and primary AWS docs.
• Use the FAQ to confirm scope, difficulty, and the services that matter most.
• Turn weak rows from this page into one-line rules and drill them under time.

Quiz