Table of Contents
Two weeks ago, if someone had asked me about most AWS AI services, I probably would have smiled confidently… and changed the topic.
Apart from a basic understanding of AWS and a few familiar services like Amazon S3, AWS IAM and Amazon Cognito, I was definitely not what you would call an AWS certification machine. In fact, this was my very first AWS certification exam.
Fast forward 14 days: I'm happy to share that I passed the AWS Certified AI Practitioner (AIF-C01) exam.
GIF via GIPHY
And no, this was not one of those heroic "I studied 9 hours a day with three monitors and motivational background music" stories.
This was a much more realistic developer story:
- working full-time,
- studying only around 1–2 hours a day on average,
- skipping some days entirely,
- and repeatedly asking myself why AWS has 47 services with names that sound equally important.
But somehow… it worked.
So I want to share exactly how I did it using completely free resources, in case you're planning to take this exam without donating your wallet to expensive courses.
The moment I opened the syllabus and questioned my life choices
When I first looked at the AWS AI Practitioner exam topics, I saw terms like:
- Amazon Bedrock
- Amazon SageMaker
- Amazon Kendra
- Amazon Comprehend
- Amazon Rekognition
- Amazon Textract
…and my brain's first reaction was:
"These all sound useful. Which is exactly the problem."
GIF via GIPHY
Because AWS exams don't test whether you've heard the service names. They test whether you know:
which service is the MOST suitable, MOST scalable, MOST cost-effective, LEAST operationally complex, and MOST AWS-ish answer.
That is a completely different sport.
A quick cheat sheet for those service names
This is the one-line mental model that helped me tell them apart:
| Service | Think of it as… |
|---|---|
| Amazon Bedrock | Use ready-made foundation models through an API to build generative AI apps |
| Amazon SageMaker | Build, train and deploy your own machine learning models |
| Amazon Kendra | Intelligent search across your organization's documents |
| Amazon Comprehend | Natural language processing: sentiment, entities and key phrases in text |
| Amazon Rekognition | Analyze images and videos: objects, faces, labels |
| Amazon Textract | Pull text, forms and tables out of scanned documents |
My study plan was not fancy. It was just sustainable.
I did not create a color-coded Notion board.
I did not print 600 pages of notes.
I did not disappear into a mountain cave.
I simply committed to:
doing at least a little every day when possible.
- Some days I studied 2 hours.
- Some days, 1 hour.
- Some days my study plan was just opening a YouTube tab, staring at it, and deciding that tomorrow is also a valid academic concept.
GIF via GIPHY
But because the resources were focused, the small daily progress stacked up surprisingly fast.
That was the key lesson:
Consistency beats dramatic one-day motivation.
The free resources that carried me through this exam
I genuinely owe this pass to free educators who put high-quality content online. No gatekeeping. No overpriced miracle courses. Just solid teaching.
GIF via GIPHY
Before anything else, it's worth bookmarking the official exam page so you know what you're signing up for:
1. My main foundation playlist on YouTube
This playlist became my starting point for understanding the entire syllabus in a simple way:
AWS AI Practitioner Full Free Playlist https://www.youtube.com/watch?v=LkBbdycWSso&list=PLgCcPWUJXIdVmjrc_I-DTrA2-LL2ff1zh
If you only pick one resource to build your foundation, make it a single structured course like this, and finish it before jumping around.
2. CloudNinja free practice tests (extremely useful)
Practice tests are where the "MOST suitable answer" style of question really clicks. Reading about a service is one thing; picking it out of four plausible options is another.
3. YouTube mock exam videos
My routine became:
- Pause the video.
- Answer the question.
- Realize I am confidently wrong.
- Learn why.
- Continue.
Very character building.
GIF via GIPHY
My unexpected MVP: ChatGPT
I have to give a special mention here.
Throughout this preparation, ChatGPT became my unofficial:
- tutor,
- doubt clearer,
- note simplifier,
- quiz generator,
- and occasionally therapist.
GIF via GIPHY
Whenever I got stuck on the classic "these two sound the same" pairs, I would just keep asking until it explained things in painfully simple English. And that mattered a lot.
Here's the short version of the pairs that tripped me up the most:
| Pair | The difference in one line |
|---|---|
| Bedrock vs SageMaker | Bedrock: use existing foundation models via an API. SageMaker: build, train and deploy your own models. |
| RAG vs fine-tuning | RAG: fetch relevant data at question time, model unchanged. Fine-tuning: train the model further on your data, so the model itself changes. |
| CloudWatch vs CloudTrail | CloudWatch: metrics, logs and alarms (how is it running?). CloudTrail: a record of API calls (who did what?). |
| Bias vs variance | High bias: the model is too simple and underfits. High variance: it memorizes the training data and overfits. |
| Embeddings vs tokenization | Tokenization: split text into tokens. Embeddings: turn data into vectors that capture meaning. |
| Governance terminology | Learn the vocabulary (responsible AI, transparency, explainability, compliance) well enough to match it to scenarios. |
If you use an AI assistant this way, ask it to quiz you on scenarios, not just to explain. That's much closer to what the exam actually does.
What actually helped me pass
Looking back, three things made the biggest difference:
1. I focused on understanding, not memorizing
AWS questions are scenario-based. If you know why a service exists, you can usually eliminate the wrong options even when the wording is unfamiliar.
2. I practiced enough scenario questions
Practice tests and mock exam videos trained me to spot keywords like "least operational overhead" or "most cost-effective" and map them to the right service.
3. I used only a few solid resources instead of 25 random ones
One course, one set of practice tests, a couple of mock exams and an AI tutor for doubts. That was it.
The best part? This cost me exactly $0
Everything I used to prepare was freely available online. Which is honestly encouraging, because it proves that:
Good guidance matters more than expensive material.
Wrapping up
This certification is special to me because:
- it's my first AWS certification,
- it was done within a short time frame,
- and it reminded me that disciplined small effort still wins.
Key takeaways
- Small, consistent sessions work. Around 1–2 hours on average, with some days skipped, was enough for me.
- Learn what each service is for. The exam rewards picking the most suitable option, not reciting names.
- Practice questions are non-negotiable. That's where the real learning happened.
- Fewer, better resources. Free, focused material beat a pile of random content.
Huge thanks to every free content creator who shares knowledge online. You are helping more careers than you realize.
And a special thank you to ChatGPT… for surviving two weeks of me asking the same AWS questions in 14 different confused formats.
GIF via GIPHY
If you're planning to take the AWS AI Practitioner exam, feel free to reach out. I'm happy to share the free roadmap and tips.
Originally published on dev.to.
