My Why
This isn't my first AWS exam, but it's the first time I challenged myself to be an early adopter/beta exam taker for any kind of technology. I'm normally one to value having exam preparation neatly packaged in Udemy format, but this one was different. I initially thought I chose this exam because of the popularity of the topic, but I quickly realized the volume of content required a bigger reason to stick to this commitment. I committed because my brain absolutely loved learning about what was possible. Studying for this exam unlocked new levels of creative problem-solving I know I'll be using professionally and personally.
When the exam was first announced, I immediately signed up for the SkillBuilder learning path. I have a license through my job and that helped solidify the decision to commit. I'd already been studying for the AWS Certified Solution Architect - Professional exam, looking to renew my AWS Certified Security - Specialty credential, and deep diving into services like Bedrock, Bedrock Agentcore, Sagemaker and fundamentals of Generative AI for work-related tasks. So while taking the exam was a stretch goal, I had a reasonable time (~6 months) of laying the foundation at a pace that I could handle while balancing my well-being, my job, and family of 5.
The day registration opened for the exam, I signed up. Self-doubt felt like it was turned up 1000 decibles once I submitted payment. But to be honest, I needed to prove something to not only myself, but to show my kids that, you can press the mute button on self-doubt through diligent preparation and courage.
How I studied
I used the SkillBuilder Exam prep plan to get familiar with all of the topics that would be covered in the exam. That learning plan include Simulearn labs that helped re-enforce much of the content that was discussed in any given domain. Not all of the domains had labs at the time I was studying, but that's likely to change once the exam transition to being generally available. Also that learning plan included sample exam question banks for each domain burried in the review and practice section for each domain. Those were incredibly helpful in familiarizing myself with the way the questions would be framed. The question bank was on par with the questions I encountered on the real exam too. The challenges with this studying tool was literally just that it was so early that some parts didn't work right or show content appropriately. There were a few typos and answers to some questions that left me feeling a little confused on why that would be the choice.
I also used Udemy course from Frank Kane and Stephane Maarek to make sure my fundamentals of machine learning and artificial intelligence capabilities were where they needed to be for AWS. The depth of content on this particular course helped close some gaps in knowledge because I hadn't been hands-on involved in the implementation of models at work. I got the opportunity to walk through some of the exercises in my own AWS account as they were reviewed through this course.
And to round it out, like I mentioned I'd already been studying for the SA - Pro and the new Security - Specialty exams as well so when it came to understanding architectural patterns, enterprise scale, security, and governance in the realm of generative AI, I felt reasonably comfortable with those concepts. For studying the SA - Pro content I used Stephane Maarek's course and for the Security - Specialty exam, I used Zeal Vora's course
Content Review
From a content perspective, the exam covered Amazon Bedrock, Agentcore, Sagemaker pretty deeply. Familiarity with the AWS Foundation Models and what modality they support is also key in this exam. Understanding the importance of Knowledge bases to support relevant findings, how to improve relevance ranking of results, vector storage requirements are also massively covered on the exam. I found that understanding "the who" the questions are focused on is just as important as understanding "the what" they are trying to achieve. The experience level of "the who" impacts which answer would be the most relevant for the question. Would you tell a business partner to deploy their own Sagemaker Endpoint to support chatbot features they want to deploy, or would Amazon Q for Business be a good fit especially if they're looking for managed integrations to a CRM solution for example.
Other topics like securing data, monitoring, logging, mitigating risks of bias and poor performance are also topics heavily covered in the exam. These topics are where I found the most cross over between my study for other exams and this one. These topics/Domain 1 and 6 were the ones I felt most confident with going into the exam.
The exam is a long one at 85 questions and 3 hours and 25 minutes so not only is it a test of knowledge, but one of endurance as well. I strongly suggest if you are going to take this beta version, stay hydrated, rest, and eat well for at least the two days prior.
What's next
I plan to continue with AWS SkillBuilder because there's part of the learning path I want to double click on because the use cases can be applied to my reality. Whether that's preparing for my kids to have "homeschool" for the summer or preparing to deploy MCP's at work, the content covered in the exam prep was just the spark I needed to bring theory to life.
Until next post folks, thanks for reading!
Top comments (0)