I'm thinking of enrolling in Georgia Tech's OMSCS program. Is it worth it for an experienced FAANG software engineer? If you've taken this program or other online masters, what's your experience like and would you recommend it to someone with industry experience? Cost-wise it looks like an incredible bargain to me: $2.2k out of pocket (with Google covering 2/3 costs) for a US masters degree. Nice to boost potential future H1B chances - i'm in a non US office so it could come handy one day. Probably going to choose ML apecialization - interesting topic to me personally, very hyped today, might help me in transitioning to an ML team
Online or not...the worth of a master's degree depends totally on the content and the specialization. Look at the courses and the course format. If you think you will learn something then for that price it is worth it. Otherwise you can spend that time in a better way by learning on your own.
Never worth it... just Leetcode and get the higher paying job. If it’s management you’re getting into, an MS isn’t going to help. MS is just another way to waste your life and pay universities money, UNLESS you are legitimately in love with a certain topic
I'm in a non-US country where Google is the only big FAANG around. Getting into another better paying FAANG requires me to pass H1B lottery and that's where a US masters degree comes in handy - it gives me higher chances due to additional masters cap ML topic seems kind of interesting to me, wouldn't mind putting some time into it tbh, degree or not. Although as I understand ML MSc are not taken too seriously, everyone wants to see at least a PhD to be considered for serious ML work
Those who voted - took omscs but don't recommend. Can you explain?
All their course material is already free on Udacity. Granted you won't get a degree, but I don't see much value addition in the OMS course. I personally believe I learnt much more on the Coursera course from Andrew Ng than OMS CS.
But courses dont get you a cs degree. Only knowledge doesn't get you interview calls.
In the program, and am enjoying it so far!
How much time commitment?
Roughly 10-20 hrs per week. Could be less if you’re familiar with the material already.
In the program, I'm not a FAANG engineer tho. Tbh it's only valuable as a line on your resume or for you, H1B leverage. The program itself is great, I just don't think master's degrees in general make you a better SWE. A better computer scientist maybe. I did a non-CS eng undergrad and find the difficulty and amount of coursework to be about 80% of what I expected a master's level CS class to be. That said it's still really hard to do those classes while working full time. ML track is probably more relevant to SWE jobs than my focus, systems
How many classes do you take per semester and how many hours a week do you spend on them total?
I'd recommend if you actually want to study at grad level rather than hoping to gain much in transferable knowledge. That being said for ML I found it too focused on applied skills. Especially classes like ML4T are basically intro to numpy
I have taken the OMSCS program. Think it was not worth it. Had multiple issues: First of all I would never get the course of my choice would always be anxious about get approved after being waitlisted. Second of all though the assignments were intense, interesting and amazing to work with, they were intensive time consuming and energy sapping. Considering that much of it was math based stuff, I was having my doubts on how often I ll use those in practice. Was great for foundation of basics though. Lastly call me bad at time management, but I used to feel very tired from work on weekdays and pretty much had only weekends to work with. Sometimes that wasn’t enough and got pretty crazy, where I had sleepless saturdays and Sundays. That got me tired on weekdays eventually. Considering all of this, i don’t think it’s worth it to transition into a new gig. Especially for a good software developer working for a great company. What I would recommend you to do first is try self paced learning of stuff through ad-hoc courses on coursera and self motivated daily but incremental learning. If that works, you steer what you want to learn and don’t need OMSCS. Bachelors or masters doesn’t matter if you have YOE. Even for H1B, I140 beyond 6 years is cap exempt. So don’t need masters for increasing odds in H1-B lottery. I blew a couple of thousands of dollars because of lack of ability to consistently self learn. Hope you don’t repeat it and make an informed decision. Good luck!
2 questions: How many classes did you take per semester? I have heard 1 is fine but 2 can lead to high stress like you described Do you think its worth it for someone with a BS CS from no-name state school? I feel like it would give me a better brand on my resume that would lead to more interviews. Im a contractor right now so i have no good "brand"
For first question, yeah I took 2. But if you take 1 course per sem might take almost 4 years. That it’s not worth it. For second, I have BS CS from a no name state college (although it was one of the best colleges in Karnataka state, India). I feel work ex in good companies, your efforts to get good exposure to quality work, and networking completely outweighs degree and where you get it. Just making sure you are in the right team doing good work, learning cutting edge stuff, constantly evolving and giving yourself opportunities to grow it should all be fine. No one even cares where I got my degree from.
Not recommended. I've joined the program but have been unable to take their Machine Learning course as it's always over subscribed. Too many people doing it. Also it's too academic - it won't help you solve real world ML problems. I found that most classes are led by TAs not professors. Do you want to spend 20 hours a week learning ML from 21 year olds? Kaggle is probably better.
Financial cost is trivial, it's really just about whether it's worth the time.
Lol @ online
My understanding is that their degree will be indistinguishable from an offline master - it's not going to state how it was completed
If barrier to entry/completion is too low, then the degree itself is diluted