L4 SDE here with the hope of becoming a Principal at FAANG or equivalent status at a startup. For those unaware, at AMZN there are only 5 SDE levels (L4-8): Principal (L7) and Senior Principal (L8). There is also the hidden level (L10) which is for Distinguished Engineers. From what I've seen, in delivering a product, a Principal SDE might be required to design, develop, test and deploy software around many technical and non-technical constraints such as: legacy dependencies, data quality, financial and legal compliance, technology limitations, pre-existing system architecture, API flexibility (i.e. the extent to which clients can build new features on top of your proposed APIs), latency requirements, scalability, design review and feature delivery timelines, SDE availability and expected operational excellence (OE) burdens. In summary, I would say that at AMZN, promotion to Senior/Principal SDE requires the ability to navigate often conflicting technical and team/organizational constraints in service of product delivery while demonstrating significant technical depth/innovation in the software you build to realize this product. At AMZN, the most important thing here is that everything is anchored by product delivery - a natural consequence of AMZN's ethos of customer obsession. So the three key skills (IMO) required for Principal (at AMZN and I assume other FAANGs) are: - Product delivery - Navigation of technical and organizational constraints - Technical depth/innovation I think things are slightly different at Google compared to AMZN: Google is an engineer's paradise where technical depth seems to be valued much more highly than at AMZN (or so I've heard). While demonstration of the ability to navigate significant technical and organizational constraints can only come through professional work experience, technical depth and algorithmic innovation is something that can be honed during PhD study. So for a recent new-grad such as myself, it seems to me that the surest path to Principal is to involve yourself in the technical areas that are bound to experience a product explosion in the near future (e.g. self-driving, personalized medicine, improved conversational agents, etc.) and leverage your technical depth in these areas to work on projects that offer opportunities for significant technical innovation in service of product delivery in the midst of nebulous technical and non-technical constraints. And I think a PhD in the "correct" subject will provide me with the technical depth needed to get an edge in engineering roles for one of these "future-product-explosion" areas (e.g. see this Waymo ML engineer ad that asks for an MS/PhD: https://waymo.com/joinus/2229501/), or potentially even launch and exit my own startup. When I look at the profiles of present-day Google Principals, a lot of them have PhDs, which seems a natural consequence of Google's high value of technical depth. At AMZN, it seems that the focus here is more heavily weighted toward product delivery than technical depth. ---- Question here ---- Basically, I'd like to get a gut-check that my intuitions are correct. To present day Principals and Senior Principals at FAANGs and hot startups, does my path to (Senior) Principal make sense? ---- Address common anti-PhD criticisms ---- I work in the UK and I've calculated the monetary opportunity cost of my PhD to be about £90k ($115k) per year over the 4-6 years of study. This is based on my present SDE-I TC shown at the bottom of this post. It's worth noting that I'm not going to have children, so this opportunity cost is not very relevant to me. And I'm willing to sacrifice this TC if it gives me greater opportunity to work on problems of greater technical depth - particularly those that will underpin the technical product explosion areas of the near future (e.g. self-driving, personalized medicine, computational drug discovery, etc.) -------- TC: £58k GBP ($74k USD) YOE: 1 #tech #phd #principal
Wow. What an offensive comment. Why am I a “waste of resources” just because I refuse to have children?
My employer is paying for my advance degree if I like. It’s always worth it to advance yourself. So what if Amazon want it or not—Amazon is good today but do you really see yourself with Amazon in 10-20 years. I am applying for part time PhD bc my employer is paying and I don’t have to worry about income lost from not working. Plus, I want to go into specific govt-field that requires higher degree in tech. See whatever works for you.
Where did you find part time phd?
Look into GW for part time PhD System Engineering and Doctor Of Engineering for Engineering Management. I got into the latter—it’s part time and I’m just having fun learning solid background Data Science courses along with Business Admin courses as well. I am not aiming to be an ML or researcher but my goal is just to gain the practical knowledge —since DEng is applied research. I can use it in the field. If you care about rankings then it might not be for you. It’s bc the uni is local and my employer is paying so it’s a win win situation for me. I’m not shelling any money and I can use what I learn at work. Plus, it’s onsite and the program tracks you for 3-yrs, so I can finish quickly. It won’t matter in tech much but in govt and defense sector, it definitely helps. The east coast is kind of old school.
Don't think you need it like in waymo. If you can pump out papers like a phd then you are as good. I feel principal is more about organization, pushing ppl, convincing ppl, be a good leader, don't think that's what you learn in a PhD.
If I'm being honest probably not. PE is a political position that requires high behavioral skills. Influence, gaining trust, networking are more important than ability to solve problems or resume. Now if you have those skills, I can see a phd helping. Having a phd means you are good at publishing and getting patents and those can help with promo.
Thanks for the reply! I’m thinking about staying at for at least the next year, by which time I’ll likely have been promoted to SDE-II. Then spend five years (2022-27) doing a PhD in ML at a top university thereafter - I attended a top 10 global university for undergrad and luckily still have the connections needed for recommendation letters. I know people that reached Principal scientist within the 10 years post-PhD. I’m hoping a PhD from a top university can give me a headstart. And I think being an engineer is a young person’s game. And this is reflected very much in the software engineer demographics at many companies. Science, however, seems like a lifelong gig.
If you want to become a scientist then for sure get a phd.
A few things about a PhD from personal experience. 1. It definitely helps to gain technical analysis, work management and planning skills as you will be juggling a lot of things simultaneously. Most of the time your PI won’t spoonfeed you and you have to create your own path of progress. Which is a good skill to learn. 2. Getting a PhD will put you on a path towards being a PE but its not a 100%. If you are in big organizations or even moderately sized startups working on cutting edge stuff, there is a high chance that most of the people already there have an advanced degree and more industry experience in that domain compared to you being a fresh graduate after PhD. 3. As you correctly pointed out, you do have an option to either begin with an early startup or start something of your own. If being a PE or working at that level early on is a goal, those might be a good start. At the end of the day it is all about experience and what you bring to the table. Your timeline of 10 years post PhD towards a PE could work in certain orgs but might be tricky in bigger ones like GOOG, MSFT, FB. Overall do a PhD only if you are ready for an experience similar to being dropped in the middle of the desert with no clear path on where to head next. (Most of the times its like that) (And don’t get me wrong, that experience had helped me immensely)
You mentioned PhD in ML - if you wanna enter Principal level as a research scientist or applied scientist, then it will certainly help. I have a PhD in phys, done a career switch to ML, and just signed my new contract as an ML applied scientist with AMZN. I feel that my PhD experience taught me how to "do science": like quickly reading through volumes of scientific papers, assessing the most promising directions of further research, performing statistical analysis or results, and publishing my own articles. PhD is certainly not a requirement to progress along any career path outside academia, but it just makes some directions easier to pursue.
Would you be willing to share your rough TC for applied scientist at Amazon?
We don’t know what the market will be for ML PhDs when you graduate from a PhD. You might also have to do a postdoc btw
[Moved to your reply on @albertosson comment]
India
5h
389
'Hindutva': The Radical Hindu Ideology That Seeks to 'Push Christianity Out of India’
Tech Industry
3d
68533
Crossed a line with my boss
AMA
Yesterday
957
PM Manager, early 40s, married and ENM (Ethical Non Monogamous) AMA
Health & Wellness
Yesterday
787
Lasik cost
Tech Industry
Yesterday
3470
What happens when most of your team is Indian?
Nice research and good question. Someone please answer! I can’t answer it.