The original plan was February 26th. Today is March 2nd, and I just closed the exam window with a "Pass" on the screen. I had to take an extra week because I realized I had underestimated the fundamentals.
I thought I could breeze through them, but the deeper I dug, the more
I realized that "knowing of" a service is not the same as understanding its architectural purpose.
The Reality Check on Study Material
In my first log, I mentioned using Whizlabs. To be honest: I wasn't happy with it this time. The questions felt slightly off-target for the AI-900.
I had to pivot my strategy:
Microsoft Practice Assessment: This was the real game-changer. It was much more aligned with the actual exam logic and helped me identify the gaps I had previously overlooked.
NotebookLM: I used this heavily to "interrogate" specific sections of the documentation. It’s one thing to read a paragraph; it’s another to have a focused tool help you deconstruct complex concepts through targeted Q&A.
The Bigger Picture: Breadth vs. Depth
The AI-102 remains my North Star because it aligns my private interests with my current professional environment. However, passing the AI-900 made me realize that I want to be more than just an "Azure implementer". I want to be a well-rounded AI Engineer. To achieve that, I’m considering broadening my horizon by working through DataCamp’s AI Fundamentals. While Microsoft gives me the tools, I want to ensure my foundation is platform-agnostic and theoretically sound.
What’s next?
The "Fundamentals" badge is on the digital shelf. For now, the "RAG-in-a-box" project is going on the back burner. I have some pressing job-related tasks that require my full attention, which will take some time. Once the dust settles, I’ll dive into the DataCamp curriculum. From there, I’ll reassess my path and decide on the next concrete steps toward becoming an AI Engineer.
Cheers,
Pat
Top comments (0)