During my journey to share my powerlifting form analysis solution with my friends, I realized that I needed to learn how to build things in the cloud. AWS has the greatest adoption and was also where my company infrasture lived, so it was a logical choice for me. I knew nearly nothing about AWS at this point, so I decided to start at the very beginning and build up from there. In the end, I spent about 5 months of my evenings and weekends picking up three different certifications that seemed potentially useful to me: Certified Cloud Practioner, Solutions Architect (Associate), and Machine Learning (Specialty). Here is a brief reflection and review of my experience.

Certified Cloud Practioner: I opted to gain this certification mostly out of fear. I knew that AWS had a tremendously wide offering of services, and I was intimidated to think about any of the other certifications without a firm foundational overview. The comprehensive free material available directly through AWS was a major selling point to me. The exam was straightforward, since it was more an exercise of memorizing than understanding or architecting. I also felt like I achieved my goal of obtaining a broad overview of lots of different services through studying for the exam. In the end, though, I would be tempted to skip this certification if I were to restart my education from scratch. I am not convinced that the actionable knowledge for someone looking to build on AWS was worth the time investment required to memorize everything needed for the exam.

Solutions Architect (Associate): I had a strong belief that studying for this certification would give me the footing I needed to start building full systems on AWS. Although I went through the free training again, I was hesitant to jump into the exam since I did not actually possess much on-prem or cloud experience. So I first completed the popular Stephane Maarek Udemy course. I enjoyed this course and felt like it adequately prepared me for the exam, although I would have had to think very hard about enrolling without an 83% discount off the sticker price. I felt like I received a solid return on my time investment studying for the course. I gained an understanding of key services deep enough to be useful for understanding design trade-offs and picked up some nice insights around system design best practices, too. In general, I would recommend this certification to anyone interested in using AWS.

Machine Learning (Specialty): I convinced myself that as an applied scientist at a company using SageMaker on AWS every day, I should add this certification just for completeness in case I could learn something useful for my job. I also knew it would be quite easy given my machine learning background and fresh solutions architect knowledge. There was also the thought in the back of my mind that potentially this would demonstrate some level of understanding to future employers. The high cost of the exam led me to again purchase a heavily discounted Udemy course to ensure that I was getting everything possible from the process of studying for the certification. I don’t have any issues with the course, since I learned everything needed for the exam, but I would skip this certification altogether if I were to start over. This was certification was the most expensive and yet I feel like I learned the least in preparing for the exam. The machine learning theory was extremely basic, and most of SageMaker is straightforward to use by design.