Leveraging AI to Prep for Optimizely Certification Exams
That time has come where my Optimizely CMS and Commerce certifications are nearing expiration. That means its time to prep for the next round of exams.
In the past my approach was straight forward but tedious. I would go through the documentation. I would consider what people recommended online. And I would review any reference material provided by Optimizely.
Being an OMVP and working with Optimizely/EPiServer for 10 years, most people would think exams like these should be straight forward. However the Optimizely exams (especially commerce), are notoriously hard. They reference lots of low level items that you may not have at the top of your mind. And there is nothing worse than having multiple choice answers that all contain code that have very slight variations.
A New Approach to Studying
This round I wanted to try something different. I recently started looking into tools like Anki for other areas of study. If you aren't familiar with Anki, it lets you build flash cards with different formats. It's true power is that it uses spaced-repetition based on how well you did with each card. So it will automatically know when to review cards you might not have done so well on the first time around.
For other subjects, I have found Anki to be very helpful. As long as you are consistent, and you don't try to cram to much in each day, it's great for long term retention of information. I thought it would be perfect for the Optimizely exam preparation.
Getting the Information
So now that I have a tool for studying, I need information to populate my flashcards. Optimizely provides reference material for all their exams. Its nicely organized into separate tables and categories, and links to documentation. Here is one table from the document.
The problem is that there are about 100 of these links. Going through each one, pulling out all the relevant data, and creating flash cards is a daunting task.
I wanted to see if I can find a smarter way of parsing all the data.
At first I thought it would be great if ChatGPT could summarize each page for me. But I was afraid too much information would be lost. Then I thought I could ask it to extract any factual data from the page. That seemed to work fairly well.
Then I started creating questions and answers in Anki based on that data. But I soon realized I could save a step and ask ChatGPT to handle that part as well.
So my process went like the following:
- Go to ChatGPT, and enter the following into the prompt
"Extract factual data from the following text and create question and answer style quiz questions and answers from it:"
- Open one of the links provided in the reference doc. I highlight and copy all the relevant text from that page.
- Go back to ChatGPT. Under the text you already typed into the prompt, paste the text you copied from the reference doc.
- Hit enter and let ChatGPT generate your questions and answers.
- Then, you can use the Anki GUI to copy and paste the questions and answers.
And boom! You have flash cards based on the Optimizely exam.
Now obviously not everything ChatGPT provided is a winner. Some items I had to omit. Other times I had to regenerate what is provided to try to get more questions. And for some reason in some instances I got questions/answers while others I got multiple choice. But I got more useful practice questions than not.
But overall it was a great way to get data into Anki fast for studying.
I am still prepping for the exam. I will post an update once I get my results. But I think overall this was a fun (and hopefully effective) way to prep for the exam. I doubt I will be able to memorize every card. But hopefully enough relevant information will stick to supplement what's already there.
And a disclaimer to anyone thinking that this is a quick win to taking the exam. I would advise more in depth studying if you haven't worked with CMS or Commerce previously. There is no substitute for having been in the code base.
But if you have some experience, and feel like you need more review on the less day to day items, this may be a good approach.
Let me know what you think. I will report back how well this worked for me. If I do well, I can also share the Anki deck I built to give others a jump start.