the microsoft dp-100 experience

originally created on 2025-07-11

updated on 2025-08-12

tags: [ms-dp100-series]
- - -

preface: this is the first series blog on my website. if a tag has "series" in it, that means that there are multiple blogs that are hidden under that tag. you can access the hidden blogs by clicking the tag in the blog list (bryanchan.org/blog/<tag>). for example, here is the tag for this series.

(the actual architecture behind it is super scuffed, but it works...feel free to check it out on my github repo.)

this summer, i had the opportunity to go to the insure+e program in normal, illinois. i made a blog about it [here]. this blog is removed. the link does not work.

however, this was only part of my summer (most of june). now, im back at home and continuing to work on what we started. however, i need to do more if i want to get a job in this market. thus, i wanted certifications.

i looked online at dozens of applications (and applied to most of them), and...not gonna lie, i got rejected. a lot. however, what i also noticed is that a lot of data scientist jobs required knowledge in cloud computing. in particular, i saw a lot of aws and azure requirements (and a bit of gcp as well). so...i decided to prove that i knew this stuff. i have done my fair share of projects, but they are all incomplete and not published. so, whats the next best way to prove that i know this stuff? certifications.

aight, lets do this.

step 0: background

i wrote this step after passing the exam. i thought it would be important to put here.

before i dive into the details, i should note that i have somewhat extensive coursework and research experience in the fields of machine learning, natural language processing and retrieval augmented generation. the reason i am stating this is because my notes also reflect this. there may be something things that i didnt cover in my notes because they were "redundant" in my head. just wanted to put discretion out there.

because of this, my notes may also seem lackluster. im not someone that really...takes notes to begin with. its still a skill that i am struggling with, so i apologize if the notes really coherent.

step 1: signing up

after some thinking, i decided to go with the microsoft dp-100 certification to start (as you can see in the title). "azure data scientist associate". not a bad title!

wow, it even has a fancy logo on it or something (source)

the first thing i did before anything else was sign up for a testing date. i am someone that tends to procrastinate a lot if i dont have a deadline, so i set one. my official testing date is august 12th, 2025. (note: i am using "is" because at the time of writing this, i have not taken the test yet).

the test fees were pretty intimidating, but i want to believe that it is worth it in the long run. the test costed approximately $91, and i decided to take it in person (i heard that taking it online was...quite annoying).

great! im signed up. now, i have to study.

step 2: study plan

so...im not really good at studying. that makes things hard...

first off, i looked at the self-paced learning path. microsoft seemed to estimate that it would take 18 hours to complete. i have a month, and i am completely unempl*yed and free (aside from working on the insure+e publication and polishing my website). so, i decided to contribute 2 hours a day to looking at this course.

after searching for valuable resources online (doomscrolling on reddit), i found a few other places to study.

i plan to document my entire study notebook in a blog series (finally using the tags for something useful hehe)

ill keep hyperlinking the sections as i go along, so you can check out blog by blog!

step 3: the study series

(update 08/12: i...lowkey...skimmed the last two modules of the self-learning course. thus, there are no notes for them...)
self-learning course module 1.1
self-learning course module 1.2
self-learning course module 1.3

more updates 08/12!

to be honest, you can kinda see the notes getting more lazily written as the modules go on. i was getting tired of writing notes, and it seemed redundant in my head. however, i should emphasize that the last module was particularly important, as it covered a lot of topics regarding azure's ai foundry - which is on the exam - and rag systems.

step 4: taking the test

i have not taken the test yet, as it is set for august 12th. i will update this section and keep writing beyond here once i take the test! :)

update 08/12: i have taken the test. i passed!! yay!!

the test itself was not too difficult, but it also was not easy. i can confidently say that the microsoft self-learning course helped a sufficient amount. microsoft's practice exam was helpful in terms of understanding basic content, but...do not expect it to be on the same difficulty level of the actual exam. this is something that microsoft itself also states.

the exam also gave me open access to microsoft learn documentation - i was allowed to search the sdk at any time. this was incredibly helpful, and i would highly recommend spending time learning how to navigate the documentation. that being said, the exam is made purposely so that you cannot pass off of the documentation alone in the given time. make sure you know your stuff!!!

overall, the certification "arc" felt like a fun side quest. i might go for aws next, since ive been hearing a lot about how it is also very useful.

hooray!

also i really need to fix my website code. that will probably be my next blog (unless by some miracle i finish rag reserach before my own website??)

thank you so much kc for checking up and improving my code