On AWS Machine Learning Engineer (Associate) Exam

Although this is not my first associate-level AWS exam that I have taken to date, I feel like this very certification deserves some attention on my blog. A brief summary of my background: I'm a software engineer with focus on cloud infrastructure and architecture. I enjoy designing and implementing systems to solve business problems - the more complex the problem the better. My career trajectory has therefore been mostly in that direction, though more recently I found myself going much further down the stack in the Linux ecosystem and IP networking using Go. Spoiler: the skills required for systems-design for small spaces is very much similar to those you would use in the cloud. But this is discussion for another day. So back on the AWS ML exam😄.

What Do You Need to Know to Pass the Exam

In order of ease based on experience, starting from more experienced to less experienced with AWS — keyword here is 'experience', so this may not apply to you if you do not have hands-on production experience with AWS:

• If you have hands-on experience with AWS and use it frequently at your day job (I'm talking designing, provisioning and deploying services on AWS using a wide range of their services especially the well known ones), AND you have experience training, testing, deploying and/or consuming machine learning models in general and maybe also doing all the machine learning stuff on AWS, then this exam will be a walk in the park. You will likely only need little time revising AWS-specific machine learning tools, models and services.
• If you have limited experience with AWS where you only use it at your day job with only a specific service or set of services AND are familiar with machine learning concepts or at least have integrated and/or consumed machine learning models, then you would likely need to do a thorough study from the ground up, starting with doing lots of hands-on in AWS. AWS recommends this exam to those who already have at least one or two of the associate-level certifications, so it makes sense that you would need to do a study tailored to one of those associate certifications first, even if you do not intend to sit the exam for them, and then come back to study for the Machine Learning Engineer exam.
• If you have general cloud infrastructure experience not specific to AWS AND have designed, trained, deployed machine learning models, your mileage would be similar to the previous point.

Final Note

As well as taking the exam, I enjoyed preparing for it too. It was both a good learning experience and a consolidation of everything that I knew about machine learning at the time of taking the exam, but more importantly in the context of AWS. Compared to AWS SysOps Administrator and Solutions Architect (Associate), this one was fairly easy and very much focused on machine learning.

Obligatory Certification Badge 😄

Click here to see my certification.