CS career megathread / AMA

[app]

If you are already getting G interviews, then I would say the boost probably isn't there since it seems like you're good at LC. But if you're not getting interviews, then I think FB almost guarantees you an interview at G (and at FAANGMULAS/whatever the acronym is now).

i mean yeah there is some truth in correlation between g and fb interview successes.

i think for pretty much anyone in the tech industry who have worked for a long number of years, getting interview at G isn't the issue at all. almost anyone can get it from what i see. the randomness and the chaos of the process is an issue. e.g. say you've worked for 12 yrs on some technology related to developing simulator for some sort of hw platform. that code is going to likely be in c/c++ and after that length of experience you're p much a reliable engineer who can do most of the shit at g. but the interview process may have some random interviewer from a very different skillset who wouldn't care much for whatever this guy did and will have his own set of questions that he or she values. it also goes the other way, a lot of eccentric old timers at g who'd throw an unreasonable question on someone who is relatively younger or doesn't have experience with what the interviewer has.
g is very random in their interviewing, with the interviewer almost having no accountability in terms of what he writes in feedback. there is also no official 'debrief' at g. the interviewer would write or rate whatever he wishes and there would be no common discussion for him to defend or justify it. at fb, they at least have an official debrief where they're on the call together unlike g to discuss their feedback.
that said, it's only my opinion, and everyone has their own unique experiences and sure there would be people out there with a different take on it.

[wizzy]

If you are already getting G interviews, then I would say the boost probably isn't there since it seems like you're good at LC. But if you're not getting interviews, then I think FB almost guarantees you an interview at G (and at FAANGMULAS/whatever the acronym is now).

i mean yeah there is some truth in correlation between g and fb interview successes.

i think for pretty much anyone in the tech industry who have worked for a long number of years, getting interview at G isn't the issue at all. almost anyone can get it from what i see. the randomness and the chaos of the process is an issue. e.g. say you've worked for 12 yrs on some technology related to developing simulator for some sort of hw platform. that code is going to likely be in c/c++ and after that length of experience you're p much a reliable engineer who can do most of the shit at g. but the interview process may have some random interviewer from a very different skillset who wouldn't care much for whatever this guy did and will have his own set of questions that he or she values. it also goes the other way, a lot of eccentric old timers at g who'd throw an unreasonable question on someone who is relatively younger or doesn't have experience with what the interviewer has.
g is very random in their interviewing, with the interviewer almost having no accountability in terms of what he writes in feedback. there is also no official 'debrief' at g. the interviewer would write or rate whatever he wishes and there would be no common discussion for him to defend or justify it. at fb, they at least have an official debrief where they're on the call together unlike g to discuss their feedback.
that said, it's only my opinion, and everyone has their own unique experiences and sure there would be people out there with a different take on it.

There's always going to be randomness with the types of questions you get asked, so luck does play a factor, but from my understanding, they still mostly have to ask questions from a pre-approved question bank. If they ask something that's banned or that HC feels like isn't an appropriate question, then their feedback can get thrown out, and you either get a chance to do an additional interview, or they can proceed with the feedback from the other interviewers.

As long as you don't get strong no hire, you can also make it through assuming you got hire / strong hire (even a couple leaning hire?) for your others. I think that interviewers are also supposed to be calibrated and that HC can question the rating when they see the type of question asked or the code submitted since interviewers are supposed to take pictures of the whiteboard (or now in covid times, the code is just saved).

I think you can argue the merits of debrief vs. no debrief, but in both cases, an outlier interview can be corrected with an outright hire or with an additional interview to get more signal. In theory, debrief could also allow an interviewer could sway a decision for better or for worse because HC is somewhat unlikely to reject a debrief decision, while HC is supposed to consist of people removed from the decisionmaking process.

I've seen it both ways with people I know rejected at FB and offered at G or rejected at G and offered at FB, so it seems like a crapshoot at a certain point with what types of questions you happened to get. Both are going to result in a bunch of false negatives.

[wizzy]

That said, to your other point: If getting the interview isn't an issue for you, then there shouldn't be too much concern about this

i'm trying to see if it'd be easier to go to G from fb or or from my current company in future.

I was thinking that G and fb basically have a revolving door between each other with how frequently people interview and hop back and forth, but if you're already at that level of getting an auto interview, that benefit isn't really there.

So then it goes back to the question of how much you value the growth opportunities at fb and the bump in comp.

[suralin]

also getting GE+ is basically impossible unless you've been downlvled. i got 2 in a row bc initially downlvled coming from a not-great background, but if I joined FB now at E5+ I would just aim for MAs

Is this for GE+ during your first cycle or are you saying in general at the higher levels for future perf as well?

Don't you need GE+ to get a promo? Or is the idea that E5 is terminal, so most people aren't progressing beyond that anyway.

mostly first cycle but also in general

there's an internal comp group that collated some distribution stats around this, but yea it's increasingly rare at higher lvls. and there's some weird handwaving around this, but technically perf ratings and promos are *supposed* to be orthogonal—you can get promoted with a mix of MAs/EEs (the most common case actually) or conversely get GEs and not promoted. there's 2 reasons for this: 1) promo is supposed to happen when you're already performing sustainably at the next lvl, whereas perf rating is more a moment in time take on your realized impact in the last half, and 2) the expectations part of the rating kind of gradually increase before e5 and after e5 they're so high that it's not really expected to be able to beat MA (unless like mentioned before you got lucky with a project etc)

[suralin]

[bootcamp, team, etc]

ah if you're on the specialist track yea that's different re: bootcamp. if the HM isn't being super clear, I would ask to speak to the actual ICs on the team, should be much more informative. if you're still feeling unsure after that, you could even try to backchannel with engineers who have left that team

re: wlb, i mean yea when I say <40 or <50 hrs/wk i def don't mean working the entire time. but yeah if even 35 hrs sounds draining already, then it might be a rough transition at first

i don't know much about the [team] anymore but FAIR has a good reputation

[suralin]

re: interviewing, yea these types of tech companies optimize for reducing false positives, so there'll naturally be a lot of FNs when ppl encounter their "anti-loop" (http://steve-yegge.blogspot.com/2008/03 ... oogle.html)

and there's definitely a revolving door. I can't find it right now, but somebody did an analysis of linkedin data looking at which companies most lead to which other companies. it's p much what you'd expect per blind's "tiers" / prestige rankings

[wizzy]

also getting GE+ is basically impossible unless you've been downlvled. i got 2 in a row bc initially downlvled coming from a not-great background, but if I joined FB now at E5+ I would just aim for MAs

Is this for GE+ during your first cycle or are you saying in general at the higher levels for future perf as well?

Don't you need GE+ to get a promo? Or is the idea that E5 is terminal, so most people aren't progressing beyond that anyway.

mostly first cycle but also in general

there's an internal comp group that collated some distribution stats around this, but yea it's increasingly rare at higher lvls. and there's some weird handwaving around this, but technically perf ratings and promos are *supposed* to be orthogonal—you can get promoted with a mix of MAs/EEs (the most common case actually) or conversely get GEs and not promoted. there's 2 reasons for this: 1) promo is supposed to happen when you're already performing sustainably at the next lvl, whereas perf rating is more a moment in time take on your realized impact in the last half, and 2) the expectations part of the rating kind of gradually increase before e5 and after e5 they're so high that it's not really expected to be able to beat MA (unless like mentioned before you got lucky with a project etc)

This is really interesting. For promos as a lagging indicator of performing at the next level, I had always thought that GE/SEE at L was supposed to be the equivalent of MA/CME at L+1. But it makes sense if the expectations are so high that you're no longer able to do things within your control to beat MA at the higher levels and it's more right place at the right time with an impactful project.

Makes all the G complaining on blind over fb's huge multipliers for GE/RE vs. SEE/S seem kinda pointless in that context.

[app]

There's always going to be randomness with the types of questions...

yeah but this all sounds good in theory about the question bank, calibration etc, it's doesn't happen so in practice. question bank is only a guideline. g is notorious for lack of accountability among its interviewers.

[app]

the 7 hr of actual work actually requires one to be really wired in for much longer. during weekday for 3-4 hrs after work still gets 'mindshare' and mind will be buzzing with work-topics until into night.

[app]

could asking to speak to ICs on the team be seen as negatively, like someone not excited enough about the group or somewhat skeptical of the group? generally only very secure managers don't see that as something negative and there can be insecure managers/people anywhere in these companies.

also any idea about

how long an offer remains valid or one can get the same offer without re-interviewing loop?

say if one says no to them, would they be able to get the same offer in Aug 2021?

or if one agrees to accept the offer, can they begin in 1 yr from now say Aug/sep 2021?

do you have to sign the offer to accept it, or one just verbally accepts it and no other paperwork before day-1?

FAIR or AML doesn't seem part of the group i mentioned, they're different BU from org perspective.

[suralin]

offers are generally valid for 1 year, and you can delay the start date for quite a while, but probably not for that long

[vonrus1]

As the law school admissions cycle kicks into full gear, it's become painfully apparent that my preferred job (120-140k, low COL area, standard bennies) just isn't feasible in law, at least not for someone right out of an average school. Sure, there may be a few here or there, but not many. This realization has made me reconsider what I want to do when I get out of the military this August.

Are jobs like that fairly attainable for CS grads from a large public university? What would be the focuses that would get there? I'm starting from a pretty basic level of knowledge. I'm seriously considering applying for a second bachelor's in CS.

[pancakes3]

As the law school admissions cycle kicks into full gear, it's become painfully apparent that my preferred job (120-140k, low COL area, standard bennies) just isn't feasible in law, at least not for someone right out of an average school. Sure, there may be a few here or there, but not many. This realization has made me reconsider what I want to do when I get out of the military this August.

Are jobs like that fairly attainable for CS grads from a large public university? What would be the focuses that would get there? I'm starting from a pretty basic level of knowledge. I'm seriously considering applying for a second bachelor's in CS.

sorry, no CS advice here, and sorry for not chiming in on your initial "chance me" thread.

is your cycle not come out the way you expected? i thought your stats with vet bump put you in the running for T13 admission. you had said JAG, which is a reasonable outcome. but i'm not going to talk you out of CS either.

i'll defer to others but i don't think going for a BS in CS is the play here; you should be thinking coding bootcamp.

[vonrus1]

As the law school admissions cycle kicks into full gear, it's become painfully apparent that my preferred job (120-140k, low COL area, standard bennies) just isn't feasible in law, at least not for someone right out of an average school. Sure, there may be a few here or there, but not many. This realization has made me reconsider what I want to do when I get out of the military this August.

Are jobs like that fairly attainable for CS grads from a large public university? What would be the focuses that would get there? I'm starting from a pretty basic level of knowledge. I'm seriously considering applying for a second bachelor's in CS.

sorry, no CS advice here, and sorry for not chiming in on your initial "chance me" thread.

is your cycle not come out the way you expected? i thought your stats with vet bump put you in the running for T13 admission. you had said JAG, which is a reasonable outcome. but i'm not going to talk you out of CS either.

i'll defer to others but i don't think going for a BS in CS is the play here; you should be thinking coding bootcamp.

Hello! Unfortunately, my stats were already unlikely for T13 even with the vet bump. JAG is also insanely competitive, and with the economic uncertainty in the next couple of years it will likely even be more competitive. Increase in apps and higher LSAT-Flex scores is a perfect storm (in a bad way) for my numbers.

I'm more concerned with being able to reliably find something like I mention above, regardless what I choose to study after I get out. Most of the law schools I'll be able to get into have so-so shots of that in a good economy, really not so much in a shakey one.

[app]

re: bootcamp. if the HM isn't being super clear, I would ask to speak to the actual ICs on the team, should be much more informative. if you're still feeling unsure after that, you could even try to backchannel with engineers who have left that team

during bootcamp, typically how many times/sessions does one talk with a given HM that one is considering joining before saying yes? do they sometimes talk multiple times (say 3-5) say 30 min each time over the bootcamp duration before agreeing to join? do they also get to talk with ICs a few different times during bootcamp before deciding?
wondering if asking for 1 chat w IC and 1 more with HM could be seen as excessive.

[jeff chiles]

I don’t know computers but an entry level job in any industry earning 120-140k in a low cost of living area is not going to be easy to find. But something at that level when you are 4-5 years in will be much easier to find (still not easy though).

[vonrus1]

I don’t know computers but an entry level job in any industry earning 120-140k in a low cost of living area is not going to be easy to find. But something at that level when you are 4-5 years in will be much easier to find (still not easy though).

Sorry, should have specified. Low COL as in not SF/SF or NYC. Thinking places like Chicago, Dallas/Houston/Austin, Colorado, Philly/Pittsburgh, Columbus. Smaller (besides Chicago) metros. Maybe I'm still overshooting it, but $120k out of school doesn't seem that crazy?

[pancakes3]

my mom was a coder and she early retired back in 2014 earning less than 120k. she had a masters and had been working in CS for 20 years at that point. tertiary market (her team was Pittsburgh/Richmond/St. Louis)

[vonrus1]

my mom was a coder and she early retired back in 2014 earning less than 120k. she had a masters and had been working in CS for 20 years at that point. tertiary market (her team was Pittsburgh/Richmond/St. Louis)

20 year career with an advanced degree, making less than 120k? Yikes. Looks like I'm way off here.

[Nony]

I mean I know full professors who’ve been teaching for 20 years who make less than $120k. (Obviously not coding but just for context.)

[vonrus1]

I mean I know full professors who’ve been teaching for 20 years who make less than $120k. (Obviously not coding but just for context.)

I guess my perception of the compensation for new SWEs or similar is skewed.

[suralin]

yeah i think you’re pretty off. i made 60k something in raleigh/durham as a new grad (c/o ‘14) with a CS degree. that number can grow quickly if you get your foot in the door at a “top” tech company (i make 10x that now) but i know lots of ex-coworkers / ex-classmates who still make <100k

120-220k is starting only if you’re talking about NYC or SF and gotten an offer from companies like google/FB/etc

[suralin]

re: bootcamp. if the HM isn't being super clear, I would ask to speak to the actual ICs on the team, should be much more informative. if you're still feeling unsure after that, you could even try to backchannel with engineers who have left that team

during bootcamp, typically how many times/sessions does one talk with a given HM that one is considering joining before saying yes? do they sometimes talk multiple times (say 3-5) say 30 min each time over the bootcamp duration before agreeing to join? do they also get to talk with ICs a few different times during bootcamp before deciding?
wondering if asking for 1 chat w IC and 1 more with HM could be seen as excessive.

i mean typically during bootcamp you’d actually sit with the team in question and work on tasks with them etc, so it’s much more organic than scheduling these 1:1s. but no i don’t think it’s excessive if that’s what you need for your decision

[icechicken]

I mean I know full professors who’ve been teaching for 20 years who make less than $120k. (Obviously not coding but just for context.)

I guess my perception of the compensation for new SWEs or similar is skewed.

Public perception is skewed by the high end (FAANG salaries, people who played the equity lottery at startups). $120k is pretty comfortable in a city like Pittsburgh and there's basically no supply pressure on employers to raise salaries in a market like that.

[wizzy]

I mean I know full professors who’ve been teaching for 20 years who make less than $120k. (Obviously not coding but just for context.)

I guess my perception of the compensation for new SWEs or similar is skewed.

You're thinking of the biglaw equivalent for SWE probably. FB and Google are paying 22 year old CS grads 200k+, but that isn't true for your run-of-the-mill SWE at a rando company in flyover territory. I know new grads in Austin making $90k at IBM (and not like IBM is a rando company either obviously).

Basically:

relatedly i was talking with somebody else about this recently and i think tech is becoming increasingly bimodal (similar to law?)

once you've got your foot in the door at a big tech company you have it made and almost certainly won't struggle to find more high-paying jobs. the whole demand >>> supply that people think applies to everybody in CS does actually apply for you

but if you don't get the right break, it's maybe surprisingly hard to get that first entry-lvl job. and with big tech increasingly hiring contractors even for 'skilled' work (there are two on my team who are new grads with CS degrees), the disparity grows. to be clear the path dependence / chicken & egg problem exists for every career but the huge pay and opportunity disparity def doesn't

e: tho still think that it's probably one of the highest ROI paths you can take if you have the potential and the interview process is relatively 'gameable'

e2: also it's relatively easy to go from not-so-well-paying to well-paying, doesn't seem like that's as possible in law