Also, don't forget about the cost of living. With 140k per year salary, about 10 percent of that (for me) was money that didn't go to taxes or bills. I have 15 years of experience (and I look young).
It's definitely a different kind of experience, so if you're feeling adventurous, do it! Just don't expect it to be fun, and be sure to do your own homework on whether there is a realistic market for a company's product if it's a startup (especially if it is Pre-Series-B). And, for the love of all that is good, don't work for any CEO under age 25. They are probably not the next Mark Zuckerberg.
Practice basic algorithms. The guys at these firms aren't nearly as smart as they think they are so you can comfortably stop after part III of the "Introduction to Algorithms" (the R. Rivest one) book (which is about 1/3 of the actual book, but like half the text is exercises, so in total this is like maybe 150 normal pages of theory, which you can condense down to 3 really dense ones, which is a good exercise by itself. Make sure to have actually written those data structures in C++ or Java (Python if you must), with tests, and if you want to ace, with benchmarks. Of course you could super-plus ace and get to part 6 or so)
And yes, the "real" way to learn software engineering is not a good way to ace these interviews. The author and maintainer of 10 years of a VM system you guys all know about, who codes and designs circles around the best engineer you've ever heard of failed the big G interview (also good enough that using some connections managed to get that reversed and an offer extended anyway). Like most jobs, it's not about actual ability, it's about knowing the "magical" bit of theory that doesn't matter in practice, but is effectively a lock on the door.
Every last bit of it you can find extensive explanations on youtube for. Don't get distracted :). There's also a free coursera course. If you don't want to follow the time schema on coursera, download the course videos on piratebay.
Oh and for academic CS candidates. Please do remember: Stanford is a LISP school, not a Haskell school. If in any book you're training for these interviews with you see the words "dependant typing", you immediately slam the book shut and make SURE you don't open it again until after the interview. Sad, but you use that stuff, your interviewer won't understand and you'll get dinged. Yes, really. Don't go into theory that you can tell your interviewer doesn't know about beyond "oh yes, I know about some stuff relating to x". DO NOT talk for 40 minutes about advanced theory your interviewer doesn't understand.
Oh and one last thing. Give a small amount of thought to how you appear to others. You might be the best engineer on the planet, but if your interviewers all write that on the feedback form, they'll pass on you. You may have 10 years of experience and get interviewed by a 2 months hire (usually 2 years, but 2 months can happen). If you proceed to shout at the guy about how his person being there disrespects you ... things won't go well. Be friendly, be approachable. Talk about your cat, but for 5 minutes at most (not for 50).
Also, you can get a small plus next to your name by asking to get referred.
I have never worked like an “engineer”. My brain takes information from many channels, I let it congeal, then I create solutions.
This means that I’m terrible at algorithms and patterns, but highly creative with real-world solutions.
I believe companies need both kinds of technical people (drone and creative), but I’ve only been in a few places that actually understand that.
But recall: $1 USD (SFO) != $1 USD (Almost anywhere else).
And there are some other downsides, you work very long hours in a company of 70%+ men. I think quality of life in South Bay for most isn't as nice as much less affluent places. It depends on your cup of tea, I guess.
https://www.nerdwallet.com/cost-of-living-calculator/compare...
That's essentially a parallel move if you consider nerdwallet doesn't consider local tax rates (very high in SFO vs 0% in Texas).
And you're also ignoring that CoL calculators bias against high CoL areas in a few important ways.
Why is it a downside for someone who has no bias against men? Would you call a downside working for example as a registered nurse just because there are too many women?
Also for comp, keep in mind that he is getting that sort of comp with two years of experience - I have met people with even less experience getting similar numbers (I know one with less experience with non-technical degree making even more). If you have the right experience/background, you can get such compensation. Just imagine what the comp is like when these people get experience under their belt at the big companies.
These companies want to hire for trajectory. If someone's gotten to a great level of skills 2 years out of university, that says good things about where they'll be in 5 or 10 years.
And don't let me get started ranting about why the majority of white collar workers even need to come in to work as often as we do just for appearances.
Silicon Valley salaries are pretty decent and I’ve heard of new grads getting 100k signing bonuses at Facebook.
Google, for example, offered a base salary just over 100k. Is the difference between level 3 and 4 (what I assume 2 years of experience gets you) really ~60%? The annual RSU difference is also drastic, as OPs offer was 100k in stock annually.
Note: I'm calculating the 60% by looking at annual salary + stock + target bonus each year.
Total comp for a new-grad Googler who doesn't negotiate at all will be 110K base +18K bonus + 25K stock annually = ~153. That's without negotiating at all. Negotiation can push that across 200K annually in extreme cases. From that perspective, its not as gargantuan a leap. And similarly, I expect that this person was able to negotiate their google offer, so it would be higher than the "default" for an L4.
I edited it to 60% to account for negotiations (assumed +10-15k annually), but I didn't think there was that much room at the new grad level. I remember Google being very inflexible on base, and didn't think you could get 200k annually with just negotiating (which would involve tripling the initial stock grant you listed, assuming that base is inflexible).
Even still, are the bands truly that wide across L3 and L4? I suppose I'm most astonished that the median L3 and top of L4 are that wide apart. Thank you for all the information.
And as to your second question: a rule of thumb that I think exists is that total comp approximately doubles every 2 levels, so you get a ~40% increase per level on the median. Keep in mind that a new grad offer is not the median of L3, but the bottom.
Taxes: https://imgur.com/a/s3SVY (Not sure if imgur links are allowed here, but I couldn't find a website that would let me calculate an estimated paycheck and share the link)
Rent: https://sfbay.craigslist.org/d/apts-housing-for-rent/search/...
Car Insurance: https://imgur.com/a/3UgPw
Public Transit Options (https://www.reddit.com/r/bayarea/comments/49dfm6/poor_public...)