AWS recently released two new exams, the AWS Certified AI Practitioner and the AWS Certified Machine Learning Engineer — Associate. This post covers the AI Practitioner certification after I recently took and passed the exam in the beta period.
The exam is marketed as being foundational, however anyone who takes it will soon learn that not all foundational exams are exam and that the AI Practitioner exam tests depth of knowledge more deeply than the Cloud Practitioner exam. Simply playing AWS Service ‘bingo’ isn’t enough to pass the AI Practitioner and to me the level of examination of subject matter seemed more inline with the level tested in the SA, Developer or SysOps Admin Associate level exams.
Most AWS exams test AWS specifics, the AWS Certified Developer Associate tests your knowledge of the services AWS offer to developers, it doesn’t test generic knowledge. AWS have taken a different approach to the AI Practitioner, in a similar way to the approach taken for the Machine Learning — Specialty certification, namely that it tests AI/ML concepts that have nothing to do with the AWS services, in addition to testing knowledge of the in-scope AWS services. So be prepared to answer general questions about why a model might be over or under fit, or what altering the temperature of a model will do, whilst also understanding that Amazon Bedrock guardrails can be used to prevent an LLM returning inappropriate content. General foundational level knowledge of AWS is also tested, for example understanding of Identity and Access Management (IAM) to limit access to resources and VPC endpoints for private network connectivity. Exam content can be reviewed here — https://d1.awsstatic.com/training-and-certification/docs-ai-practitioner/AWS-Certified-AI-Practitioner_Exam-Guide.pdf.
As the exam is currently in beta, the material available for studying is limited. I utilised the AWS Skill Builder Standard Exam Prep Course, which is listed as 8 hours of content, as well as the ‘Official Practice Question Set’, which gives access to 20 questions. Whilst I passed the exam I didn’t feel this course alone would be enough for someone without a good existing knowledge of AWS and wider AI/ML knowledge I’d acquired taking the AWS Machine Learning — Specialty certification. I’ve not got a lot of experience of Skill Builder courses but I also found it difficult to consume as it was very ‘dry’, versus the style of my preferred platform, ACloudGuru. There is a ‘Standard Exam Prep Plan’ course, which is listed as 18 hours of content, which does look like it covers the general AI/ML topics on top (8 hours of this course is the Standard Exam Prep Course). If you have access to an AWS Skill Builder paid subscription then you’d have the option of the ‘Enhanced Exam Prep Course’ (16.5 hours) or the ‘Enhanced Exam Prep Plan’, although this comes in at 29 hours of content, but it does look a lot more comprehensive. A paid subscription also opens up access to the full practice exam and a 2 hour AWS Escape Room exam prep. The subscription is $29/month for pay as you go, so you could limit costs to $29 if you’re disciplined and have plenty of study time available.
The only other options I’ve seen at the time of writing are Udemy courses, which are currently ‘on sale’ (when aren’t they!). However, looking at the current content listed it appears you need to be very careful about which course you purchase and there are a variety of options
The course by Stephane Maarek is 9 hours of content and includes a practice exam, it’s currently available for £14.99. It covers a lot of topics in a short amount of time so you may need to go over some of them a couple of times. It does cover both the general concepts and the AWS services that are tested in the exam. The also another course of similar length by Sundog Education by Frank Kane that is also priced at £14.99 and appears to cover the general concepts and AWS services. There are other options too that either have fewer reviews, lower overall scores, or both.
The course by Rasool Shaik appears to be an investment in the hope the course material will be added to over time, worrying given it wa one of my top Google search results for an exam prep course for this certification. If you’re looking to study for the exam imminently I’d steer well clear as it’s only 1 hour long and lacks covering an AI / ML topics and seems nothing more than a placeholder to me.
The exam is listed as having a number of different exam question styles, some of them which are entirely new, to me at least, they’re listed in the exam guide: -
Ordering and Matching are both new formats, whereas case study is really a variation on existing themes, it’s simply that several questions asked related to the same scenario presented, rather than having separate scenarios per question.
Example ‘ordering’ question from the free practice questions: -
Example ‘matching’ question from the free practice questions: -
As with all questions it’s really important to fully read the question and understand what it’s asking, for example where a question asks for lowest latency, or highest availability.
In summary the AWS Certified AI Practitioner course and exam covers some interesting topics for those starting out on their AI and AWS AI journey, providing a grounding in both generic techniques and AWS specific services including Bedrock and Amazon Q from a GenAI perspective, as well as more traditional ML services. However just because the cert is badged as ‘foundational’ it shouldn’t be underestimated, nor compared to being of the same level as Cloud Practitioner. If you’ve already got a good grounding of AWS AI ML services and general concepts I’m sure it’ll be easy enough, but if you haven’t I would definitely recommend a decent amount of study before attempting.
