SteadfastIQ

SteadfastIQ Blog

AWS AI Practitioner study guide: a first certification in AI on AWS

About a 5-minute read

AWS Certified AI Practitioner (exam code AIF-C01) is a foundational exam. It is meant to show that you understand how machine learning and generative AI are used on AWS, including when they are a bad idea, not that you can already build a production training pipeline. That makes it a strong first credential if your work is starting to touch AI services, or if you want a structured way into the topic before an associate-level exam.

This guide is a study plan for the AWS AI Practitioner full course on SteadfastIQ: objective videos, hands-on labs, flashcards, practice questions, and a study guide. SteadfastIQ is not affiliated with, endorsed by, or sponsored by Amazon Web Services. Confirm every exam-format detail on AWS’s official page when you register.

Exam format and domains, as currently published

According to AWS’s current public information and our course overview, AIF-C01 is a 90-minute exam with 65 questions (50 scored and 15 unscored). Question types include multiple choice, multiple response, ordering, and matching. The published passing score is 700 on a scaled 100–1,000 range. Always recheck those numbers before you book.

The current domain weights are:

  1. Fundamentals of AI and ML (20%)
  2. Fundamentals of Generative AI (24%)
  3. Applications of Foundation Models (28%)
  4. Guidelines for Responsible AI (14%)
  5. Security, Compliance, and Governance for AI Solutions (14%)

Two domains are easy to under-study because they sound “soft”: Responsible AI, and Security, Compliance, and Governance. Together they are 28% of the exam — the same weight as Applications of Foundation Models. A plan that only covers how models work is an incomplete plan.

Who this exam is for (and who should pick a different first cert)

This is a good first exam if you can already navigate basic cloud ideas or you are willing to learn them as you go, and you want the AI-specific vocabulary of AWS: what a foundation model is, how a managed service is used, what a prompt or a retrieval pattern is for, and what guardrails exist for.

If you have never used AWS at all, consider spending a week in the AWS Cloud Practitioner course first, or at least its fundamentals domain, so accounts, shared responsibility, and billing are not brand-new on the same day as tokens and embeddings. If you already build models for a living, this exam may feel introductory; the AWS Machine Learning Engineer – Associate course may be a better match, or a practitioner pass may still be useful if your employer wants a common baseline.

If your workplace is Azure-first, look at Azure AI Fundamentals rather than forcing an AWS logo. The AWS or Azure first-cloud guide has a fuller decision rule.

A five-week plan that follows the weights

Week 1: Fundamentals of AI and ML. Supervised versus unsupervised learning, basic training and inference ideas, when a classical ML approach is enough, and the AWS services the current outline expects you to recognize at a practitioner level. Use videos and the study guide, then a lab that makes you choose a path, not only name a service. Flashcards for terms you cannot yet explain to a colleague in one sentence.

Week 2: Fundamentals of Generative AI. What generative models do, their limits, and the operational ideas around them (cost, latency, hallucination as a risk, and the difference between using a model and training one). Stay practical. If a concept cannot be tied to a decision — “we should use this because…” or “we should not use this because…” — you do not have it yet.

Week 3: Applications of Foundation Models. This is the heaviest domain. Retrieval patterns, prompts, customization options as described in the current objectives, and how AWS expects you to select a service for a use case. Do every hands-on lab in this domain. Practice questions should be mixed, not only from the chapter you just read, so you have to select an approach.

Week 4: Responsible AI plus Security, Compliance, and Governance. Treat these as one serious week, not a Friday afternoon. Responsible AI includes bias, transparency, human oversight, and use policies — in the form the official objectives use. Security includes access, data protection, and governance of AI services on AWS, including the shared-responsibility ideas that also appear in other AWS exams. Labs and scenario questions matter more than slogans.

Week 5: Mixed practice and repair. Take timed sets in the course, including items that use ordering or matching if the bank includes those types, so the format is not a surprise. Review every miss against the official domain. Rewatch only the videos that map to clusters of misses. Re-run the labs you would not want to explain in an interview. Then book or confirm your official exam date.

How to study services without memorizing the catalog

AWS has many AI-related services, and the outline does not ask you to become a walking price list. Learn services in clusters: a service you would use to call a foundation model, a service you would use to build a broader application, a service you would use for more traditional ML, and the security or governance tools that wrap them. For each cluster, be able to answer: what problem it is for, what it is not for, and one risk of using it carelessly.

When two answers in a practice question both look “AI-ish,” the better one is usually the one that matches the use case constraints: data sensitivity, cost, latency, or the need to stay within a documented policy. That is practitioner judgment, and it is also how week 4 leaks into week 3.

Labs, the range, and what not to expect

Hands-on labs in the course are the place to lock those use-case choices. They are mobile compatible, so a shorter lab can happen on a phone; use a desk when you are clicking through a console path you have not seen before.

Hack the Globe is a cybersecurity live-fire range, not an AWS AI lab. It is still worth a session if you want practice thinking about abuse, logging, and control — themes that show up in the governance domain — but it will not replace AIF-C01 labs. Do not hide there from a service-selection exercise.

This course, and this exam, are about using and governing AI services. They are not a claim about how any particular website is built. Study the official AWS objectives and the course in front of you.

Start with the official outline and one lab

Open the AWS AI Practitioner course and the official exam page. Read domain 1 of the study guide, watch the first video cluster, and do the first lab on the same day. Five weeks later you should be able to explain, in ordinary language, when a generative model helps, when it does not, and how you would keep the work inside a security and responsibility boundary. That is the exam, and it is also the job skill the badge is supposed to represent.

Open the coursePlay Hack the GlobeAll certifications