AWS AI Practitioner Study Plan: A 4-Week Guide to AIF-C01

The AWS Certified AI Practitioner (AIF-C01) is a foundational exam with 65 questions in 90 minutes, and a pass mark of 700 on a 100 to 1000 scale. A four-week plan works well for most beginners: week one for AI and ML basics, week two for generative AI and foundation models, week three for applying them, and week four for responsible AI, security and mock exams.

One caution before you start. Exam details and domain weightings change, so confirm the current format, domains and objectives in the official AWS exam guide. Treat everything below as a structure to hang that guide on, not a replacement for it.

Step zero: read the exam guide

Before you study anything, download the official exam guide and read it twice. It tells you which domains are tested, how they are weighted, and what the exam expects you to be able to do in each. It is the only reliable statement of what is in scope.

Turn it into a checklist. Every objective in the guide becomes a line you can tick off, and any line you cannot explain in a sentence or two is a gap. You will come back to this checklist every week.

The 4-week plan at a glance

Week	Theme	Main goal
1	AI and machine learning fundamentals	Learn the vocabulary and the ML lifecycle
2	Generative AI and foundation models	Understand what they are, how they work and their limits
3	Applications of foundation models	Prompting, retrieval-augmented generation, customisation, evaluation
4	Responsible AI, security and governance, then mocks	Close gaps and rehearse the real exam

A realistic rhythm is 30 to 60 minutes on most weekdays plus one longer weekend session. Adjust it to your life; consistency matters more than intensity.

Week 1: AI and ML fundamentals

Start with the language. Foundational exams reward people who can tell similar terms apart.

• How AI, machine learning and deep learning relate to each other.
• Supervised, unsupervised and reinforcement learning, with one example of each.
• The typical ML lifecycle: data, training, evaluation, deployment, monitoring.
• Common problem types, such as classification, regression, clustering and recommendation.
• Basic ideas about data quality, overfitting and why evaluation metrics matter.
• Where AWS offers AI and ML services, at the level of “what is this service for”. Learn the purpose of each, and leave the configuration details alone.

End the week by writing, from memory, a one-line definition of every term you met. Turn the ones you could not do into flashcards.

Week 2: Generative AI and foundation models

This is the heart of the exam, so give it time.

• What a foundation model is and how it differs from a model trained for one task.
• Large language models and other model types at a conceptual level.
• Tokens, context and why model output is probabilistic.
• Typical use cases, such as summarisation, chat, code assistance and image generation.
• Strengths and limits, including hallucination and why outputs need checking.
• How AWS makes foundation models available, and the general idea of choosing a model for a task. Keep your understanding at a general level and confirm specifics in the official documentation.

A good habit this week: for every concept, write down one place it helps and one place it fails.

Week 3: Applications of foundation models

This week is about using models well, which is where scenario questions tend to live.

• Prompting. Clear instructions, examples in the prompt, and how wording changes output.
• Retrieval-augmented generation (RAG). The idea of giving a model relevant documents at question time, and why that helps with current or private information.
• Customisation options. The general difference between prompting, retrieval, and fine-tuning or further training, and when each might make sense.
• Evaluating output. How you would judge whether a generative application is working.
• Matching the tool to the job. Many questions describe a business need and ask which approach fits. Practise spotting the simplest option that meets it.

Do your first short set of practice questions at the end of this week, even if you feel unready. The point is to see the style, not to score well.

Week 4: Responsible AI, security, governance and mock exams

Responsible AI

Expect questions on fairness, bias, transparency, explainability and human oversight. Learn what each means in plain terms and how a team might address it in practice, for example by reviewing training data or keeping a person in the loop for high-stakes decisions.

Security and governance

Cover the general principles: protecting data used with AI systems, controlling who can access what, and keeping a record of how systems are used. Stay at the level the exam guide describes. Foundational questions are more about understanding the shared responsibility idea than configuring anything.

Mock exams

Spend the back half of the week on full-length practice runs under exam conditions: 65 questions, 90 minutes, no notes. Afterwards, spend at least as long reviewing as you did answering. Sort every miss into one of three groups: did not know it, confused two concepts, or misread the question. Each group has a different fix. Our post on how to use practice exams goes into this, and the last two weeks before an exam post is a good companion for the final stretch.

How to study each week

Reading alone will not carry you through, and the same retrieval-based approach applies here as for any exam. Each session:

1. Spend the first part learning a new topic from your course or notes.
2. Spend the rest recalling it: closed-book summaries, flashcards or practice questions. This is the difference between active recall and rereading.
3. Start the next session with a quick review of the previous one.

Flashcards are well suited to the vocabulary-heavy parts of this exam, and spaced repetition means you review each card just before you would forget it. Lernoa’s flashcards are free and unlimited, with an SM-2-style scheduler, so you can build a deck as you go. Any tool that schedules reviews would do the job.

Using Lernoa for AIF-C01, honestly

Lernoa has a catalog of free certification prep courses, which you can work through alongside the official guide. There is also a timed exam simulator for each supported certification. The simulator is the one paid part: you unlock it once per certification with credits, and it is then yours to use as often as you like within fair-use limits. It has two modes: practice mode works through the whole question bank without a timer, and exam mode is timed and weighted by domain.

You do not need it to pass. Plenty of people use AWS’s own materials, free practice questions and a good notes system. The simulator is a convenient option if you want a realistic rehearsal and would rather not assemble one yourself. See pricing for how credits work, and check which certifications currently have a simulator on the certifications page.

For the wider picture on preparing for any certification, see our post on passing a certification exam on the first try.

Next step

Download the current exam guide today and turn its objectives into a checklist. Then book week one in your calendar and start with the fundamentals. If you want a place to build your vocabulary deck as you go, you can start free and look through the free prep courses in the catalog.