AWS AIP-C01 Study Plan: Bedrock, RAG, and Guardrails in 30, 60, and 90 Days

Use this study plan when you want a structured route through AWS Certified Generative AI Developer - Professional (AIP-C01). This is not a prompt-vocabulary exam. AIP-C01 tests whether you can build, integrate, secure, evaluate, operate, and troubleshoot production GenAI applications on AWS.

AWS’s current exam guide frames AIP-C01 around foundation model integration, data management, compliance, implementation, integration, safety, governance, optimization, testing, validation, and troubleshooting. The strongest answers usually protect the data boundary, ground the answer, constrain tool use, evaluate behavior, monitor cost and quality, and keep rollback or human review available where the scenario requires it.

Current exam facts

I verified these current AWS exam facts on May 16, 2026.

AWS’s certification overview lists the exam as 75 questions over 180 minutes. The AIP-C01 exam guide also lists ordering and matching item types, so practice sequence and pairing logic explicitly instead of preparing only for ordinary single-answer questions.

Weight-driven study allocation

The first two domains are 57% of scored content, but safety and testing are not optional. In real AIP-C01 scenarios, the best answer usually crosses domains: RAG plus privacy, agents plus tool permissions, streaming plus latency, or evaluation plus rollback.

Pick the right timeline

Choose the longer route if you still confuse Bedrock model invocation, Knowledge Bases, Agents, Guardrails, vector stores, IAM/KMS boundaries, evaluation, monitoring, and workflow orchestration as one generic “AI app” bucket.

The AIP-C01 study loop

Use one production GenAI loop every week.

    flowchart LR
	  U["Use case"] --> D["Data boundary"]
	  D --> M["Model and pattern"]
	  M --> I["Integration path"]
	  I --> C["Controls"]
	  C --> E["Evaluation"]
	  E --> O["Operate and optimize"]
	  O --> R["Miss rule"]

Minimum production baseline

You do not need a large lab, but you should be able to explain these paths without guessing:

30-day intensive plan

Use this route only if you already have strong AWS application experience and basic GenAI implementation exposure.

30-day rule

Every study day should create one artifact.

60-day balanced plan

This is the best default route for most candidates.

90-day part-time plan

Use this route if you are building production GenAI depth while studying.

What to drill by domain

High-yield GenAI comparisons

Ordering and matching practice

AIP-C01 can include ordering and matching items. Practice these explicitly.

For ordering and matching, partial familiarity is not enough. You need the exact sequence or exact pairing.

Final-week checklist

Use the final week for decision speed, not new topic sprawl.

Readiness signals

You are close to ready when:

• You can explain why each wrong answer violates a data, integration, safety, evaluation, cost, or operational constraint.
• You can separate RAG failures from prompt failures, model-fit failures, tool failures, guardrail failures, permission failures, and runtime failures.
• You can choose controls for PII, prompt injection, unsafe outputs, overprivileged agents, missing audit evidence, and hallucination risk.
• You can evaluate a GenAI app using measurable criteria instead of subjective “quality” language.
• You can answer ordering and matching items without treating them like ordinary multiple-choice questions.
• You can keep a steady pace across 75 questions in 180 minutes.

If you only have 48 hours

This is not ideal for a professional-level exam, but if you are already near-ready:

1. Read the cheat sheet twice: once before practice and once after reviewing misses.
2. Drill one mixed timed block and classify every miss by domain and failure layer.
3. Spend one focused block on RAG/data boundaries and one on agents/safety/security/governance.
4. Review operations and testing traps: cost, latency, monitoring, evaluation, rollback, ordering, matching, and troubleshooting order.
5. Recheck the current official AWS page and exam guide before scheduling or buying an attempt.

Booking signal

Schedule only when your misses are narrow and explainable. If you still choose answers by familiar AI vocabulary instead of use case, data boundary, architecture pattern, safety control, evaluation evidence, or operational constraint, keep studying. AIP-C01 rewards production GenAI judgment under professional-level time pressure.