CS career megathread / AMA

guest

Re: CS career megathread / AMA

Post by guest » Fri Sep 14, 2018 8:40 pm

suralin wrote:
Mon Sep 10, 2018 11:17 pm
guest wrote:
Sun Sep 09, 2018 6:55 pm
this is great. how much of the 50 LC problems are medium and hards?

i have used the listed github sys design primer, but i feel sys design interview is very subjective and unpredictable even after it's done.
about 20 were medium, v few hard. i generally feel like hards are a waste of time in terms of EV so i just skim the discussion for those / look up explanations. but airbnb gave me a LC hard in a phone interview so uh. it was palindrome pairs; i did ok bc i had seen it before but not good enough apparently

yeah it's subjective but that doesn't have to be a bad thing. gives you a lot of room to dive deep on a specific design decision or tradeoff area. eg i enjoy talking about the CAP theorem
guest wrote:
Sun Sep 09, 2018 7:35 pm
that's a big difference in TC at FANG and the generic companies.

went through the interview process and LC in the last couple of months. got offered in ML at one of the legacy companies with an okay tc so may be can use it to get ml at fang next year.
what do you mean by legacy company? would take the offer regardless tho (unless your opportunity cost is high), can always shoot for FANG later
It's old economy company like, ICE (ibm, cisco, ebay) or amd/intel.

Onsited at FB. Per recruiter coding rounds were great. I thought sys design went well (my research area) but the interviewer was much inexperienced and that's where subjectivity comes in.
But got an offer at a legacy in ml at about your target tc range. still much less than what fb would have given if they offered.

Kodokushiest
Posts: 2725
Joined: Thu Jan 25, 2018 5:38 pm

Re: CS career megathread / AMA

Post by Kodokushiest » Fri Sep 14, 2018 9:49 pm

Every prospective law student with a CS degree should have to read this thread.

guest

Re: CS career megathread / AMA

Post by guest » Sat Sep 15, 2018 5:58 pm

wizzy wrote:
Wed Sep 12, 2018 12:50 am
I'm guessing IBM, Oracle, HP, Dell/EMC, etc?

And that's really good insight into SWE vs. DE. I knew a little bit about the PM and UX distinction as compared to SWEs from blind but had no idea about the SWE/DE split. Interviewing for TC bidding war purposes makes a ton of sense.
yes about the bolded. i'd include ebay, cisco, intel, amd because they don't apparently pay as much. i think the term ICE is also used to refer to these on forums like blind.

but i think in ML space, they have to pay *somewhat* competitively in order to compete with the likes of Apple, google. so i was able to up my tc a bit into 260k range due to ml group in one of these, but i think for regular SWE roles even after 10+ years of work ex, tc at these companies is <200k, whereas i think fb/google would pay 500-600k/yr tc for that much domain experience.

User avatar
wizzy
Posts: 5347
Joined: Thu Jan 25, 2018 2:43 pm

Re: CS career megathread / AMA

Post by wizzy » Sat Sep 15, 2018 6:48 pm

Kodokushiest wrote:
Fri Sep 14, 2018 9:49 pm
Every prospective law student with a CS degree should have to read this thread.
Every prospective law student should have to read this thread.

I still remember some of suralin's earliest posts in the lounge (I think I first talked to him in BB2 and FNS) where he was posting about maybe law school but also finishing up his CS degree after starting CS classes later in his undergraduate career and how people were like holy fuck do CS do not go to law school (which I'm sure he already knew at that point, but few people actually take that to heart), and now he's on the path to world domination at FB —or FB + GOOG or FB + unicorn.

Kodokushiest
Posts: 2725
Joined: Thu Jan 25, 2018 5:38 pm

Re: CS career megathread / AMA

Post by Kodokushiest » Sat Sep 15, 2018 7:04 pm

wizzy wrote:
Sat Sep 15, 2018 6:48 pm
Kodokushiest wrote:
Fri Sep 14, 2018 9:49 pm
Every prospective law student with a CS degree should have to read this thread.
Every prospective law student should have to read this thread.

I still remember some of suralin's earliest posts in the lounge (I think I first talked to him in BB2 and FNS) where he was posting about maybe law school but also finishing up his CS degree after starting CS classes later in his undergraduate career and how people were like holy fuck do CS do not go to law school (which I'm sure he already knew at that point, but few people actually take that to heart), and now he's on the path to world domination at FB —or FB + GOOG or FB + unicorn.
Pretty sure it will just make humanities majors who are a few years out of school sad about their life choices. But I guess there's value in that.

User avatar
wizzy
Posts: 5347
Joined: Thu Jan 25, 2018 2:43 pm

Re: CS career megathread / AMA

Post by wizzy » Sat Sep 15, 2018 7:21 pm

Kodokushiest wrote:
Sat Sep 15, 2018 7:04 pm
wizzy wrote:
Sat Sep 15, 2018 6:48 pm
Kodokushiest wrote:
Fri Sep 14, 2018 9:49 pm
Every prospective law student with a CS degree should have to read this thread.
Every prospective law student should have to read this thread.

I still remember some of suralin's earliest posts in the lounge (I think I first talked to him in BB2 and FNS) where he was posting about maybe law school but also finishing up his CS degree after starting CS classes later in his undergraduate career and how people were like holy fuck do CS do not go to law school (which I'm sure he already knew at that point, but few people actually take that to heart), and now he's on the path to world domination at FB —or FB + GOOG or FB + unicorn.
Pretty sure it will just make humanities majors who are a few years out of school sad about their life choices. But I guess there's value in that.
Becoming a self-hating lawyer is a big part of becoming a lawyer.

Kodokushiest
Posts: 2725
Joined: Thu Jan 25, 2018 5:38 pm

Re: CS career megathread / AMA

Post by Kodokushiest » Sat Sep 15, 2018 7:22 pm

wizzy wrote:
Sat Sep 15, 2018 7:21 pm
Kodokushiest wrote:
Sat Sep 15, 2018 7:04 pm
wizzy wrote:
Sat Sep 15, 2018 6:48 pm
Kodokushiest wrote:
Fri Sep 14, 2018 9:49 pm
Every prospective law student with a CS degree should have to read this thread.
Every prospective law student should have to read this thread.

I still remember some of suralin's earliest posts in the lounge (I think I first talked to him in BB2 and FNS) where he was posting about maybe law school but also finishing up his CS degree after starting CS classes later in his undergraduate career and how people were like holy fuck do CS do not go to law school (which I'm sure he already knew at that point, but few people actually take that to heart), and now he's on the path to world domination at FB —or FB + GOOG or FB + unicorn.
Pretty sure it will just make humanities majors who are a few years out of school sad about their life choices. But I guess there's value in that.
Becoming a self-hating lawyer is a big part of becoming a lawyer.
:lol:

User avatar
wizzy
Posts: 5347
Joined: Thu Jan 25, 2018 2:43 pm

Re: CS career megathread / AMA

Post by wizzy » Sat Sep 15, 2018 7:23 pm

guest wrote:
Sat Sep 15, 2018 5:58 pm
wizzy wrote:
Wed Sep 12, 2018 12:50 am
I'm guessing IBM, Oracle, HP, Dell/EMC, etc?

And that's really good insight into SWE vs. DE. I knew a little bit about the PM and UX distinction as compared to SWEs from blind but had no idea about the SWE/DE split. Interviewing for TC bidding war purposes makes a ton of sense.
yes about the bolded. i'd include ebay, cisco, intel, amd because they don't apparently pay as much. i think the term ICE is also used to refer to these on forums like blind.

but i think in ML space, they have to pay *somewhat* competitively in order to compete with the likes of Apple, google. so i was able to up my tc a bit into 260k range due to ml group in one of these, but i think for regular SWE roles even after 10+ years of work ex, tc at these companies is <200k, whereas i think fb/google would pay 500-600k/yr tc for that much domain experience.
That's awesome. How did you get into sys design/ML?

guest

Re: CS career megathread / AMA

Post by guest » Sun Sep 16, 2018 3:03 pm

wizzy wrote:
Sat Sep 15, 2018 7:23 pm
guest wrote:
Sat Sep 15, 2018 5:58 pm
wizzy wrote:
Wed Sep 12, 2018 12:50 am
I'm guessing IBM, Oracle, HP, Dell/EMC, etc?

And that's really good insight into SWE vs. DE. I knew a little bit about the PM and UX distinction as compared to SWEs from blind but had no idea about the SWE/DE split. Interviewing for TC bidding war purposes makes a ton of sense.
yes about the bolded. i'd include ebay, cisco, intel, amd because they don't apparently pay as much. i think the term ICE is also used to refer to these on forums like blind.

but i think in ML space, they have to pay *somewhat* competitively in order to compete with the likes of Apple, google. so i was able to up my tc a bit into 260k range due to ml group in one of these, but i think for regular SWE roles even after 10+ years of work ex, tc at these companies is <200k, whereas i think fb/google would pay 500-600k/yr tc for that much domain experience.
That's awesome. How did you get into sys design/ML?
i didn't specifically apply for ML or distributed sys design. in fact, i have zero experience in either. my case is not regular on a ls forum like this as i have a ug and grad degree in ece. i worked on parallel computing software in c++ for > decade or so in tech companies. i wanted to switch to law for various reasons so didn't focus much on advancing career in tech companies. prepped for lsat which i thought was very hard.

realizing i was grossly unpaid after hearing about fang comps (tc was at 180k and i didn't care before) i recently began looking around. thought fb would offer, they didn't, but unexpectedly the legacy company did in their ml division for new chipsets, as apparently ml algorithms (tensorFlow etc) have computations that are highly parallel and parallel computing swe work applies there. even though they old companies still pay low tc for their traditional openings, they invest more in ml so they offered about tc of 50% more than where i was. can't apply to FB for 1 year so now taking up the offer to make up for comp disparity seems to be better than staying at current one.

guest

Re: CS career megathread / AMA

Post by guest » Tue Sep 18, 2018 5:53 pm

suralin wrote:
Mon Sep 10, 2018 11:17 pm
about 20 were medium, v few hard. i generally feel like hards are a waste of time in terms of EV so i just skim the discussion for those / look up explanations. but airbnb gave me a LC hard in a phone interview so uh. it was palindrome pairs; i did ok bc i had seen it before but not good enough apparently

yeah it's subjective but that doesn't have to be a bad thing. gives you a lot of room to dive deep on a specific design decision or tradeoff area. eg i enjoy talking about the CAP theorem

what do you mean by legacy company? would take the offer regardless tho (unless your opportunity cost is high), can always shoot for FANG later
gluck on the sep 19th interview

let us know in detail how it goes.

User avatar
suralin
Posts: 3944
Joined: Thu Jan 25, 2018 11:36 am

Re: CS career megathread / AMA

Post by suralin » Thu Sep 20, 2018 8:09 pm

v quick update:

google onsite retro: five 45-min coding rounds, got a LC backtracking* hard that i had never seen before, did decent in another, and p well on the remaining 3 (one of which was also a hard, graph traversal this time, but i knew the pattern). really have no idea how i did overall or how high the bar is, but not super confident tbh.

in better news, i got an offer from snap and they went way over the top w ~$320k (!) not incl refreshers. sinking ship etc, but the equity vests monthly w no cliff, so minimal downside imo. also it’s the exact ML-type role i want w an impressive team. re upside/downside, actually curious about your take on it wiz
wizzy wrote:
*ugh i knew this was a weakness going in, would have preferred a dp problem

User avatar
suralin
Posts: 3944
Joined: Thu Jan 25, 2018 11:36 am

Re: CS career megathread / AMA

Post by suralin » Thu Sep 20, 2018 8:14 pm

guest wrote:
Sun Sep 16, 2018 3:03 pm
wizzy wrote:
Sat Sep 15, 2018 7:23 pm
guest wrote:
Sat Sep 15, 2018 5:58 pm
wizzy wrote:
Wed Sep 12, 2018 12:50 am
I'm guessing IBM, Oracle, HP, Dell/EMC, etc?

And that's really good insight into SWE vs. DE. I knew a little bit about the PM and UX distinction as compared to SWEs from blind but had no idea about the SWE/DE split. Interviewing for TC bidding war purposes makes a ton of sense.
yes about the bolded. i'd include ebay, cisco, intel, amd because they don't apparently pay as much. i think the term ICE is also used to refer to these on forums like blind.

but i think in ML space, they have to pay *somewhat* competitively in order to compete with the likes of Apple, google. so i was able to up my tc a bit into 260k range due to ml group in one of these, but i think for regular SWE roles even after 10+ years of work ex, tc at these companies is <200k, whereas i think fb/google would pay 500-600k/yr tc for that much domain experience.
That's awesome. How did you get into sys design/ML?
i didn't specifically apply for ML or distributed sys design. in fact, i have zero experience in either. my case is not regular on a ls forum like this as i have a ug and grad degree in ece. i worked on parallel computing software in c++ for > decade or so in tech companies. i wanted to switch to law for various reasons so didn't focus much on advancing career in tech companies. prepped for lsat which i thought was very hard.

realizing i was grossly unpaid after hearing about fang comps (tc was at 180k and i didn't care before) i recently began looking around. thought fb would offer, they didn't, but unexpectedly the legacy company did in their ml division for new chipsets, as apparently ml algorithms (tensorFlow etc) have computations that are highly parallel and parallel computing swe work applies there. even though they old companies still pay low tc for their traditional openings, they invest more in ml so they offered about tc of 50% more than where i was. can't apply to FB for 1 year so now taking up the offer to make up for comp disparity seems to be better than staying at current one.
ahh that’s a dope story. i’m glad you interviewed around / this is why comp transparency is important. grats!!

User avatar
suralin
Posts: 3944
Joined: Thu Jan 25, 2018 11:36 am

Re: CS career megathread / AMA

Post by suralin » Thu Sep 20, 2018 8:16 pm

wizzy wrote:
Sat Sep 15, 2018 6:48 pm
Kodokushiest wrote:
Fri Sep 14, 2018 9:49 pm
Every prospective law student with a CS degree should have to read this thread.
Every prospective law student should have to read this thread.

I still remember some of suralin's earliest posts in the lounge (I think I first talked to him in BB2 and FNS) where he was posting about maybe law school but also finishing up his CS degree after starting CS classes later in his undergraduate career and how people were like holy fuck do CS do not go to law school (which I'm sure he already knew at that point, but few people actually take that to heart), and now he's on the path to world domination at FB —or FB + GOOG or FB + unicorn.
haha thanks man. yeah i remember those days well. tls singlehandedly changed my mind (and life? around). i joined in nov 2012, which was the v same semester i took my first cs class ever. always have a soft spot for tls/lsl bc of it

User avatar
wizzy
Posts: 5347
Joined: Thu Jan 25, 2018 2:43 pm

Re: CS career megathread / AMA

Post by wizzy » Fri Sep 21, 2018 12:34 am

suralin wrote:
Thu Sep 20, 2018 8:09 pm
v quick update:

google onsite retro: five 45-min coding rounds, got a LC backtracking* hard that i had never seen before, did decent in another, and p well on the remaining 3 (one of which was also a hard, graph traversal this time, but i knew the pattern). really have no idea how i did overall or how high the bar is, but not super confident tbh.

in better news, i got an offer from snap and they went way over the top w ~$320k (!) not incl refreshers. sinking ship etc, but the equity vests monthly w no cliff, so minimal downside imo. also it’s the exact ML-type role i want w an impressive team. re upside/downside, actually curious about your take on it wiz
wizzy wrote:
*ugh i knew this was a weakness going in, would have preferred a dp problem
Congrats man, that’s an amazing offer and career opportunity. I would be really tempted to make the jump if I were in your shoes. Totally get the snarky sinking ship comments on Blind and elsewhere, but that kind of job seems like exactly the risk/reward play you’d want to make at this stage in your life when you’re young, have a long investment horizon, and don’t have children/family relying on you to maintain the big, stable company lifestyle (even if you did, your TC is through the roof anyway).

If you don’t mind me asking, what’s the base/bonus/on-hire breakdown (feel free to give a range or whatever)? I hesitate to say “rock bottom” because nobody really knows how far SNAP will fall, but if you’re looking at $600k-$800k in 2018 SNAP stock over 4 years, that could easily have a value of $1.2M-$1.6M averaged across vesting dates. You could definitely be looking at a situation where you walk into an almost-JD-level $500k income year in 2019 or 2020.

I really do think that SNAP will climb back to $20+ at some point but obvi have no idea when, and you seem to have a pretty safe floor with the monthly equity vesting, your in-demand skillset, and your ability to boomerang to FB or hit up GOOG/[top-tier tech company or unicorn] whenever means that you won’t have to go down with the ship if you sense things going bad.

From a valuation perspective, it’s hard to make a bull case for SNAP with their growth slowing, net income getting even more negative, poor DAU numbers, low gross margins, and FB/insta competition. They’ve fallen out of favor with Wall Street, but I don’t think I’d ever bet against a company like that in the era of ~data mining~. You obviously understand the technical side of big data better than I do, but I look at this as potentially getting in on the ground floor of SNAP—or at least joining at 2016 valuation levels. I think it’s really cool to be part of something like that even beyond the financials: hot area of tech, impressive team, could open up even more doors down the road, serious impact in helping grow a company that was a Wall Street darling not too long ago.

FB IPO (May 2012): $38
LNKD IPO (May 2011): $45
TWTR IPO (Nov 2013): $26
YELP IPO (Mar 2012): $15
SNAP IPO (June 2017): $17

FB 15 months post-IPO: $37 (-3%) [-50% at low in Aug 2012]
LNKD 15 months post-IPO: $107 (138%) [Fell 40% in Feb 2016 from $186 to $110]
TWTR 15 months post-IPO: $38 (-16%) [-50% at low in June 2016]
YELP 15 months post-IPO: $30 (100%) [$15 at low in Feb 2016]
SNAP 15 months post-IPO (i.e., right now): $9 (-47%) [all-time low]

All of those companies were hit hard at one point with respected analysts predicting the demise of their company, yet they recovered nicely. There's admittedly survivorship bias there, given that I chose social media companies that are still alive today, but I think the point still stands.

Snapchat probably has more growing pains, and there’s a legitimate bear case against it that tons of people are betting on, given short interest and the outstanding put-call ratio. Nobody is saying that SNAP will ever reach FB-level success, and I’m sure you have access to all the data showing how Whatsapp and Insta stories are murdering Snapchat in DAUs, while Snapchat’s DAU/MAU YOY numbers have also slowed. They’re still burning through their FCF at a rate of more than $200 million per quarter, they’re missing basically every revenue/gross margin/net income/user adoption target, and their costs are climbing, further eating into their bottom line (top-line growth slowing along with other metrics). Their EV/EBITDA is -10; their EV/FCF is -16.

It’s a really grim financial story, but I don’t think those metrics, outside of DAU/MAU, are how you should be valuing a tech company still in its early stages. SNAP might not be the next FB and Spiegel might not be Zuckerberg, but like Facebook pre-Sheryl Sandberg, Snapchat can still focus on their scalable business model and riding out the lack of profitability and just building out their ecosystem. There are definitely ways to better monetize their userbase—FB makes like 10x in ARPU metrics?—and they’ll eventually figure it out. I just can’t see the market remaining bearish on SNAP forever when the way tech has been headed for several years now, and seemingly will continue to be headed with the social media/code repository M&A activity, in big data + customer acquisition.

This is your chance to get JD rich—even if you end up just getting Snapchat and X company from your list in a bidding war. Speaking of which, there’s a decent chance that GOOG or at least one of the others on your list match SNAP, right?
Last edited by wizzy on Fri Sep 21, 2018 12:38 am, edited 1 time in total.

User avatar
wizzy
Posts: 5347
Joined: Thu Jan 25, 2018 2:43 pm

Re: CS career megathread / AMA

Post by wizzy » Fri Sep 21, 2018 12:36 am

* I hold no positions in any stocks mentioned except FB, which I am balls deep in.

** None of this should be taken as financial advice. And standard disclaimer that I’m a pretty big risk taker (sometimes to the point of being dumb).

*** SNAP in 2018 is probably still dicey and could continue to fall, which makes me wary in the short term, but I’d be shocked if SNAP in 2020 is still trading at $9/share.

User avatar
suralin
Posts: 3944
Joined: Thu Jan 25, 2018 11:36 am

Re: CS career megathread / AMA

Post by suralin » Fri Sep 21, 2018 1:47 am

really appreciate the super in-depth analysis <3

yeah generally strongly agree that this is exactly the kind of risk/reward play i'm in a good position to take advantage of, and also agree that the downside is relatively capped and there's potentially a ton of upside. i've been getting some skepticism from friends and family so wanted to make sure i wasn't crazy in how i'm approaching this
Congrats man, that’s an amazing offer and career opportunity. I would be really tempted to make the jump if I were in your shoes. Totally get the snarky sinking ship comments on Blind and elsewhere, but that kind of job seems like exactly the risk/reward play you’d want to make at this stage in your life when you’re young, have a long investment horizon, and don’t have children/family relying on you to maintain the big, stable company lifestyle (even if you did, your TC is through the roof anyway).

If you don’t mind me asking, what’s the base/bonus/on-hire breakdown (feel free to give a range or whatever)? I hesitate to say “rock bottom” because nobody really knows how far SNAP will fall, but if you’re looking at $600k-$800k in 2018 SNAP stock over 4 years, that could easily have a value of $1.2M-$1.6M averaged across vesting dates. You could definitely be looking at a situation where you walk into an almost-JD-level $500k income year in 2019 or 2020.

I really do think that SNAP will climb back to $20+ at some point but obvi have no idea when, and you seem to have a pretty safe floor with the monthly equity vesting, your in-demand skillset, and your ability to boomerang to FB or hit up GOOG/[top-tier tech company or unicorn] whenever means that you won’t have to go down with the ship if you sense things going bad.

From a valuation perspective, it’s hard to make a bull case for SNAP with their growth slowing, net income getting even more negative, poor DAU numbers, low gross margins, and FB/insta competition. They’ve fallen out of favor with Wall Street, but I don’t think I’d ever bet against a company like that in the era of ~data mining~. You obviously understand the technical side of big data better than I do, but I look at this as potentially getting in on the ground floor of SNAP—or at least joining at 2016 valuation levels. I think it’s really cool to be part of something like that even beyond the financials: hot area of tech, impressive team, could open up even more doors down the road, serious impact in helping grow a company that was a Wall Street darling not too long ago.
current offer breakdown is something like 160k base + 500k/4 RSUs + 25-50k annual bonus + 20k signing (haven't really negotiated but i did mention my other onsites; i think base + equity is p formulaic, tho they mentioned they're willing to move on signing)

yup, i'm not super bullish either, but could definitely see the stock recovering to at least the high teens. the important thing, as you point out, is that the no-cliff monthly vesting + ability to walk away gives me optionality / some ability to treat this like a double sided coin. i'm reminded of this article on valuing startup option packages https://www.benkuhn.net/optopt — obviously my case is lower risk (and variance) still than that. also yeah it is tempting to view this as the equivalent of getting in pre-ipo

if i look at my average job tenure, it's something like 1.7 yrs —> 0.7 —> 1.4 yrs, so i'm not too worried about not having the mindset to get out early-ish if necessary. i also think that snapchat has enough of a niche that if it does fail, it'll do so by wall st standards (ie not "completely"), and will likely still have a decent sticky plateau of engaged users and or die slowly enough that i should be able to stick it out for a yr's worth. i think it's super plausible (w some potentially-non-public analysis backing the idea [insert obligatory disclaimers]) that social apps have a sort of ratchet effect: once you get a user on your app it's hard to lose them; there are many cases where apps stop growing but don't shrink; app declines happen faster between cohorts than within; XAP growth v decline is not symmetric; and ratchet effects are neatly explained by switching (and possibly sunk) costs + network effects (crucially these networks/clusters can be at a fine granularity)

one thing i'm potentially worried about is it being seen as a black mark on my resume, but probably a misplaced concern / projecting #preftige worry. my sense in tech is that past a certain point, your actual experience is more important than company name (rep is probably most important early on as an easy way to signal that you've been prescreened), and then it probably inverts when you get senior enough to materially affect a company's fortunes. i feel good about the people i'd be working with—my barometer is the clichéd 'if you're the smartest person in a room find a new room' and i def won't be the smartest
Snapchat probably has more growing pains, and there’s a legitimate bear case against it that tons of people are betting on, given short interest and the outstanding put-call ratio. Nobody is saying that SNAP will ever reach FB-level success, and I’m sure you have access to all the data showing how Whatsapp and Insta stories are murdering Snapchat in DAUs, while Snapchat’s DAU/MAU YOY numbers have also slowed. They’re still burning through their FCF at a rate of more than $200 million per quarter, they’re missing basically every revenue/gross margin/net income/user adoption target, and their costs are climbing, further eating into their bottom line (top-line growth slowing along with other metrics). Their EV/EBITDA is -10; their EV/FCF is -16.

It’s a really grim financial story, but I don’t think those metrics, outside of DAU/MAU, are how you should be valuing a tech company still in its early stages. SNAP might not be the next FB and Spiegel might not be Zuckerberg, but like Facebook pre-Sheryl Sandberg, Snapchat can still focus on their scalable business model and riding out the lack of profitability and just building out their ecosystem. There are definitely ways to better monetize their userbase—FB makes like 10x in ARPU metrics?—and they’ll eventually figure it out. I just can’t see the market remaining bearish on SNAP forever when the way tech has been headed for several years now, and seemingly will continue to be headed with the social media/code repository M&A activity, in big data + customer acquisition.

This is your chance to get JD rich—even if you end up just getting Snapchat and X company from your list in a bidding war. Speaking of which, there’s a decent chance that GOOG or at least one of the others on your list match SNAP, right?
agreed re legit bear case and i def don't think snap will displace FB in the marketplace. yup i do believe there's a bunch of low-hanging fruit and high-level am not too pessimistic w how snap fits into overall tech trajectory

i doubt google would wholly match snap, but i think uber and twitter probably would

alright not going to decide on anything for a while but feeling better about things (not [only] bc confirmation bias, but bc i like talking out and exploring decisions like this), thanks again dude

Kodokushiest
Posts: 2725
Joined: Thu Jan 25, 2018 5:38 pm

Re: CS career megathread / AMA

Post by Kodokushiest » Fri Sep 21, 2018 8:27 am

Congratulations, suralin, on the amazing offer!

User avatar
wizzy
Posts: 5347
Joined: Thu Jan 25, 2018 2:43 pm

Re: CS career megathread / AMA

Post by wizzy » Fri Sep 21, 2018 8:47 pm

suralin wrote:
Fri Sep 21, 2018 1:47 am
really appreciate the super in-depth analysis <3

yeah generally strongly agree that this is exactly the kind of risk/reward play i'm in a good position to take advantage of, and also agree that the downside is relatively capped and there's potentially a ton of upside. i've been getting some skepticism from friends and family so wanted to make sure i wasn't crazy in how i'm approaching this
Congrats man, that’s an amazing offer and career opportunity. I would be really tempted to make the jump if I were in your shoes. Totally get the snarky sinking ship comments on Blind and elsewhere, but that kind of job seems like exactly the risk/reward play you’d want to make at this stage in your life when you’re young, have a long investment horizon, and don’t have children/family relying on you to maintain the big, stable company lifestyle (even if you did, your TC is through the roof anyway).

If you don’t mind me asking, what’s the base/bonus/on-hire breakdown (feel free to give a range or whatever)? I hesitate to say “rock bottom” because nobody really knows how far SNAP will fall, but if you’re looking at $600k-$800k in 2018 SNAP stock over 4 years, that could easily have a value of $1.2M-$1.6M averaged across vesting dates. You could definitely be looking at a situation where you walk into an almost-JD-level $500k income year in 2019 or 2020.

I really do think that SNAP will climb back to $20+ at some point but obvi have no idea when, and you seem to have a pretty safe floor with the monthly equity vesting, your in-demand skillset, and your ability to boomerang to FB or hit up GOOG/[top-tier tech company or unicorn] whenever means that you won’t have to go down with the ship if you sense things going bad.

From a valuation perspective, it’s hard to make a bull case for SNAP with their growth slowing, net income getting even more negative, poor DAU numbers, low gross margins, and FB/insta competition. They’ve fallen out of favor with Wall Street, but I don’t think I’d ever bet against a company like that in the era of ~data mining~. You obviously understand the technical side of big data better than I do, but I look at this as potentially getting in on the ground floor of SNAP—or at least joining at 2016 valuation levels. I think it’s really cool to be part of something like that even beyond the financials: hot area of tech, impressive team, could open up even more doors down the road, serious impact in helping grow a company that was a Wall Street darling not too long ago.
current offer breakdown is something like 160k base + 500k/4 RSUs + 25-50k annual bonus + 20k signing (haven't really negotiated but i did mention my other onsites; i think base + equity is p formulaic, tho they mentioned they're willing to move on signing)

yup, i'm not super bullish either, but could definitely see the stock recovering to at least the high teens. the important thing, as you point out, is that the no-cliff monthly vesting + ability to walk away gives me optionality / some ability to treat this like a double sided coin. i'm reminded of this article on valuing startup option packages https://www.benkuhn.net/optopt — obviously my case is lower risk (and variance) still than that. also yeah it is tempting to view this as the equivalent of getting in pre-ipo

if i look at my average job tenure, it's something like 1.7 yrs —> 0.7 —> 1.4 yrs, so i'm not too worried about not having the mindset to get out early-ish if necessary. i also think that snapchat has enough of a niche that if it does fail, it'll do so by wall st standards (ie not "completely"), and will likely still have a decent sticky plateau of engaged users and or die slowly enough that i should be able to stick it out for a yr's worth. i think it's super plausible (w some potentially-non-public analysis backing the idea [insert obligatory disclaimers]) that social apps have a sort of ratchet effect: once you get a user on your app it's hard to lose them; there are many cases where apps stop growing but don't shrink; app declines happen faster between cohorts than within; XAP growth v decline is not symmetric; and ratchet effects are neatly explained by switching (and possibly sunk) costs + network effects (crucially these networks/clusters can be at a fine granularity)

one thing i'm potentially worried about is it being seen as a black mark on my resume, but probably a misplaced concern / projecting #preftige worry. my sense in tech is that past a certain point, your actual experience is more important than company name (rep is probably most important early on as an easy way to signal that you've been prescreened), and then it probably inverts when you get senior enough to materially affect a company's fortunes. i feel good about the people i'd be working with—my barometer is the clichéd 'if you're the smartest person in a room find a new room' and i def won't be the smartest
Snapchat probably has more growing pains, and there’s a legitimate bear case against it that tons of people are betting on, given short interest and the outstanding put-call ratio. Nobody is saying that SNAP will ever reach FB-level success, and I’m sure you have access to all the data showing how Whatsapp and Insta stories are murdering Snapchat in DAUs, while Snapchat’s DAU/MAU YOY numbers have also slowed. They’re still burning through their FCF at a rate of more than $200 million per quarter, they’re missing basically every revenue/gross margin/net income/user adoption target, and their costs are climbing, further eating into their bottom line (top-line growth slowing along with other metrics). Their EV/EBITDA is -10; their EV/FCF is -16.

It’s a really grim financial story, but I don’t think those metrics, outside of DAU/MAU, are how you should be valuing a tech company still in its early stages. SNAP might not be the next FB and Spiegel might not be Zuckerberg, but like Facebook pre-Sheryl Sandberg, Snapchat can still focus on their scalable business model and riding out the lack of profitability and just building out their ecosystem. There are definitely ways to better monetize their userbase—FB makes like 10x in ARPU metrics?—and they’ll eventually figure it out. I just can’t see the market remaining bearish on SNAP forever when the way tech has been headed for several years now, and seemingly will continue to be headed with the social media/code repository M&A activity, in big data + customer acquisition.

This is your chance to get JD rich—even if you end up just getting Snapchat and X company from your list in a bidding war. Speaking of which, there’s a decent chance that GOOG or at least one of the others on your list match SNAP, right?
agreed re legit bear case and i def don't think snap will displace FB in the marketplace. yup i do believe there's a bunch of low-hanging fruit and high-level am not too pessimistic w how snap fits into overall tech trajectory

i doubt google would wholly match snap, but i think uber and twitter probably would

alright not going to decide on anything for a while but feeling better about things (not [only] bc confirmation bias, but bc i like talking out and exploring decisions like this), thanks again dude
That's a great article. Love the thoughts on startup valuation, and I definitely agree with the thought process. I totally get the black mark on resume and #preftige concerns (this is TLS/LSL, after all, and it's hard not to have the equivalent of Vault rankings in the back of your mind), but I would think that your resume would still look amazing with Facebook and Snapchat. SNAP seems like the equivalent of the sexy lit boutique option that isn't a V1-V2 like FB but pays above-market comp and gives you really good substantive experience in an area you're both interested in and that's on fire in tech (would assume actual experience trumps company brand beyond initial signaling, as you said, and I don't see a situation where you wouldn't have recruiters hitting you up). I know that's not the greatest analogy because lol at law, and you know way more about the technical side of the industry than I do, but I still think it's a really great opportunity with where you are in your life/career and your ability to swing for the fences here.

Skepticism from friends/family seems natural because why ever leave a FB/GOOG-tier company??? But man, those numbers are eye-popping and are only going to continue to increase for someone with your skillset. I think the upside of that $500k equity turning into $1M+ is enticing as fuck, and even if you sense things not going well, you'll still have plenty of options since you're clearly willing to put yourself through the leetcode grind to achieve your goals regardless of whether that's comp or better role. I obviously agree that the right play here is to view the offer as an option contract of sorts where you hopefully are in a golden handcuffs situation a year from now since you'll have gotten in at 2016 pre-IPO valuation levels, but if not, you will still have monthly vesting, be making 40% more than you are right now, and have the ability to get back to FB or try one of the other 20 options on your list.

At minimum, you've effectively leveled yourself up in terms of comp and role, right?

Also agree with you that failing by Wall Street standards is completely different than failing by ordinary company standards since the former deals with market cap, while the latter deals with whether it's a viable company. Tons of tech companies, past and present, were good companies but not Wall Street darlings. AMZN was flat for a 5-year period starting in 2003 when analysts dinged them for their non-existent profitability. Similar story for the first 18 months of FB during the heavy user-adoption/beginning of monetization/growth phase but before the real ad revenue started flowing in. MSFT was a literal dinosaur from 2000-13 pre-Nadella and pre-cloud. Same deal with AAPL pre-iPhone/iPad and for a short period post-Steve Jobs. I'm again only picking large-cap winners with a heavy dose of survivorship bias, but I'm a big believer in SNAP eventually being able to monetize its product just with the sheer number of users it has and the amount of data available, and if that somehow ends up not happening, then that's not really your problem, and you boomerang to FB or try out Google/Linkedin/Dropbox/Netflix/Uber/Lyft/Airbnb.

It's cool to follow the process through your mini liveblog, and it obviously takes balls to throw out all the companies you're interviewing with and talk bad interviews along with good interviews, so props for that.


*Bolded is super interesting and something that I hadn't really thought of but makes sense on an intuitive level. Falls in line with how a company can be stable and even becoming progressively more profitable even if it isn't hitting aggressive analyst benchmarks for user adoption. I could see a case where SNAP is a success without being a big-time Wall Street success, although I do think that eventually it will be both regardless of how many people are currently shorting it and how murky the outlook might look at present. Your ceiling at SNAP seems crazy high if/when they get their shit together, so I think the sinking ship narrative [as it relates to accepting an offer there, given the risk/reward potential] is somewhat lazy/contrived.

guest

Re: CS career megathread / AMA

Post by guest » Sat Sep 22, 2018 5:57 pm

suralin wrote:
Thu Sep 20, 2018 8:09 pm
v quick update:

google onsite retro: five 45-min coding rounds, got a LC backtracking* hard that i had never seen before, did decent in another, and p well on the remaining 3 (one of which was also a hard, graph traversal this time, but i knew the pattern). really have no idea how i did overall or how high the bar is, but not super confident tbh.

in better news, i got an offer from snap and they went way over the top w ~$320k (!) not incl refreshers. sinking ship etc, but the equity vests monthly w no cliff, so minimal downside imo. also it’s the exact ML-type role i want w an impressive team. re upside/downside, actually curious about your take on it wiz
wizzy wrote:
*ugh i knew this was a weakness going in, would have preferred a dp problem
ah, yeah hard problems can be a hit or miss depending on whether seen before or not. if it's LC can you share or pm? which level were you being interviewed for? (i think goog may weigh them less depending on level.)

congrats on the offer, that's amazing bump! probably more than the best out there. i somehow never submitted mine at snap. but not sure if submission would even get me an interview considering i have submitted mine at msft couple of times and it goes to a black box somehow even after decades in the industry.

guest

Re: CS career megathread / AMA

Post by guest » Sat Sep 22, 2018 6:11 pm

suralin wrote:
Thu Sep 20, 2018 8:14 pm
guest wrote:
Sun Sep 16, 2018 3:03 pm
wizzy wrote:
Sat Sep 15, 2018 7:23 pm
guest wrote:
Sat Sep 15, 2018 5:58 pm
wizzy wrote:
Wed Sep 12, 2018 12:50 am
I'm guessing IBM, Oracle, HP, Dell/EMC, etc?

And that's really good insight into SWE vs. DE. I knew a little bit about the PM and UX distinction as compared to SWEs from blind but had no idea about the SWE/DE split. Interviewing for TC bidding war purposes makes a ton of sense.
yes about the bolded. i'd include ebay, cisco, intel, amd because they don't apparently pay as much. i think the term ICE is also used to refer to these on forums like blind.

but i think in ML space, they have to pay *somewhat* competitively in order to compete with the likes of Apple, google. so i was able to up my tc a bit into 260k range due to ml group in one of these, but i think for regular SWE roles even after 10+ years of work ex, tc at these companies is <200k, whereas i think fb/google would pay 500-600k/yr tc for that much domain experience.
That's awesome. How did you get into sys design/ML?
i didn't specifically apply for ML or distributed sys design. in fact, i have zero experience in either. my case is not regular on a ls forum like this as i have a ug and grad degree in ece. i worked on parallel computing software in c++ for > decade or so in tech companies. i wanted to switch to law for various reasons so didn't focus much on advancing career in tech companies. prepped for lsat which i thought was very hard.

realizing i was grossly unpaid after hearing about fang comps (tc was at 180k and i didn't care before) i recently began looking around. thought fb would offer, they didn't, but unexpectedly the legacy company did in their ml division for new chipsets, as apparently ml algorithms (tensorFlow etc) have computations that are highly parallel and parallel computing swe work applies there. even though they old companies still pay low tc for their traditional openings, they invest more in ml so they offered about tc of 50% more than where i was. can't apply to FB for 1 year so now taking up the offer to make up for comp disparity seems to be better than staying at current one.
ahh that’s a dope story. i’m glad you interviewed around / this is why comp transparency is important. grats!!
i never would have known such high total comps existed until i looked around, and would have probably remained in a bubble. it's amazing how so many people work with advanced or graduate degrees in CS from well known colleges at well known companies for 20 plus years but in traditional/old-economy roles and are paid peanuts compared to these types of tc. it was unbelievable at first, i thought people were joking when they talked 100k or so/per year RSUs. i'm still not anywhere at that level in terms of RSUs but at least the current change is better than what i had before.

could you describe the SNAP ml-role?

User avatar
suralin
Posts: 3944
Joined: Thu Jan 25, 2018 11:36 am

Re: CS career megathread / AMA

Post by suralin » Wed Oct 03, 2018 11:13 am

closing the loop:

i signed the snap offer! recruiter was initially pushing me to join the LA team (ads targeting) and dangling a 50k signing bonus* that would not be in play if i picked SF (personalization/recommendations) — managed to get them to extend it for the latter after passing another abbreviated interview with SF and receiving a strong rec from the hiring manager. another condition was that i make a decision within 24 hours, which is annoying but not unfair since it's been >2 weeks. total comp with signing amortized over 2 years, 25% refresher grant (starting before nov 1st so i can be eligible for this round's), and assuming no discretionary bonus is ~340k.

@lastguest re: role details, i'll be doing feed ranking and story personalization and recommendations. it won't be 'pure' ML, but as close as it gets without a phd and research scientist title, so not just feature engineering and data pipeline work

@wiz: ty for the write-up, agreed with the law comparison and that the sinking ship contention is kinda lazy. also my friends have been v supportive so i think blind is heavily skewing my perception of their reputation

other offers / onsites: i also received offers from opendoor and twitter**, but neither even close to snap's TC. i debated trying to continue to negotiate, but honestly ready to be done with it all. google is somehow still in the hiring committee process, but i'm assuming eventual no hire (at best a severe lowball and downlevel) based on recruiter unresponsiveness. rejected from pinterest; the team only had 1 headcount and there was a senior DS who had his onsite the day before. had canceled onsite with stitch fix. i also had an uber onsite for next week and a lyft interview with the level 5 team (their self-driving division, like uber's ATG). i will cancel those since i am not the sort of person who would continue interviewing after accepting an offer 👀


*also in RSUs but w a 6 month vest and no clawback stip if i leave before 1 yr, which is the typical condition

**on a lighter note the twitter office and food are unexpectedly awesome. if i had to rank 'top-tier' tech company free food offerings, it'd be something like: dropbox > FB > twitter >> snap | google >> youtube > pinterest >>> amazon

TheWalrus
Posts: 365
Joined: Fri Jan 26, 2018 8:20 pm

Re: CS career megathread / AMA

Post by TheWalrus » Thu Oct 04, 2018 9:45 pm

Congrats dude. When do you start?

(Also, the Albert, Bernard, and Cheryl problem made me feel super dumb)

User avatar
suralin
Posts: 3944
Joined: Thu Jan 25, 2018 11:36 am

Re: CS career megathread / AMA

Post by suralin » Fri Oct 12, 2018 8:54 pm

ok so.. i'm glad but also in a slight dilemma:

i ended up doing the uber onsite and i think bc i had nothing to lose? i did really well. like easily the best set of interviews i've ever had and both coding problems* were completely new to me too

anyway ended up getting an uber L4 offer that's at the top of their range:

$154,350 base (maxed out salary range) + $480k RSUs over 4 yrs + $15k annual bonus + $55k annual refresher grant over 3 yrs
= ~$310k

the initial grant has a 1 year cliff, the refreshers don't—monthly vesting for both. she said they can add a sign-on bonus monday. i'm p confident that uber will be IPOing in 2019. the RSUs are benchmarked off their latest 409a valuation, not the preferred price from the latest investment

comparing uber to snap, uber wins in terms of business outlook, ~prestige~ / company rep, and probably quality of ppl i'll work w. snap has the edge on role. i'd be a software engineer, backend, at both but snap is ML and uber will be more data infra. comp is kind of a wash? the thing with snap's TC is that the RSUs are based on the average closing price of aug & sep and so the numbers have already dropped quite a bit. the thing w uber is regardless of IPO confidence it's still not real money yet. upsides are likely close?

reneging would suck but it's not a HUGE deal. (big fan of the "companies will screw you over at the drop of a hat, they're not your friend, look out for yourself first" school of thought)

idk i have a lot more to say but will leave it at that for now. opinions v welcome
wizzy wrote:would again love any thoughts you have
*in total the onsite was 2 coding rounds + 1 system design + 1 eng director behavioral + 1 'bar raiser' combination; the bar raiser is a senior person outside of the hiring team who has ~a mandate to keep overall company hiring bar high~

User avatar
wizzy
Posts: 5347
Joined: Thu Jan 25, 2018 2:43 pm

Re: CS career megathread / AMA

Post by wizzy » Fri Oct 12, 2018 9:49 pm

suralin wrote:
Fri Oct 12, 2018 8:54 pm
ok so.. i'm glad but also in a slight dilemma:

i ended up doing the uber onsite and i think bc i had nothing to lose? i did really well. like easily the best set of interviews i've ever had and both coding problems* were completely new to me too

anyway ended up getting an uber L4 offer that's at the top of their range:

$154,350 base (maxed out salary range) + $480k RSUs over 4 yrs + $15k annual bonus + $55k annual refresher grant over 3 yrs
= ~$310k

the initial grant has a 1 year cliff, the refreshers don't—monthly vesting for both. she said they can add a sign-on bonus monday. i'm p confident that uber will be IPOing in 2019. the RSUs are benchmarked off their latest 409a valuation, not the preferred price from the latest investment

comparing uber to snap, uber wins in terms of business outlook, ~prestige~ / company rep, and probably quality of ppl i'll work w. snap has the edge on role. i'd be a software engineer, backend, at both but snap is ML and uber will be more data infra. comp is kind of a wash? the thing with snap's TC is that the RSUs are based on the average closing price of aug & sep and so the numbers have already dropped quite a bit. the thing w uber is regardless of IPO confidence it's still not real money yet. upsides are likely close?

reneging would suck but it's not a HUGE deal. (big fan of the "companies will screw you over at the drop of a hat, they're not your friend, look out for yourself first" school of thought)

idk i have a lot more to say but will leave it at that for now. opinions v welcome
wizzy wrote:would again love any thoughts you have
*in total the onsite was 2 coding rounds + 1 system design + 1 eng director behavioral + 1 'bar raiser' combination; the bar raiser is a senior person outside of the hiring team who has ~a mandate to keep overall company hiring bar high~
Oh nice man, that's awesome! Congrats!!

Definitely agree with you on "companies will screw you over at the drop of a hat, they're not your friend, look out for yourself first," so beyond having an uncomfortable convo, I wouldn't really care much about reneging. I'm all for employee autonomy, and I think it's kinda baked into tech hiring anyway with the way that people hop jobs in search of a better role/higher TC over the boomer company loyalty mindset.

I also think Uber will IPO in 2019 from what I've read and from talking to people who seem to know what they're talking about. It's harder to evaluate Uber's financial outlook since they don't have public disclosure requirements, so it's more difficult to dig into the numbers that support their valuation, but they've also been one of the most anticipated IPOs of the past 5 years, and I think that they're more likely than not to at least have a successful IPO day/week and that you'd (hopefully) see some stock appreciation between now and then.

Would there be any lock-up period for you going into the IPO?

Also, what do the Snap stock numbers look like now with the average of August/September?

I think I'd lean Uber from everything you've written since they seem to check a lot of your boxes with respect to future of the business, prestige, quality of peers, etc. To me, Snap made a lot of sense vs. FB, but Uber seems to have a lot of the upsides Snap has with possibly fewer downsides.

I obviously don't know nearly as much about the technical side of things as you do, but how much do you value the Snap ML role vs. data infrastructure? Which aspect of infrastructure (if they break it down by data platform, network, dev infra, SRE)? And do you have any insight into Uber's workplace culture (or if that's just Blind gossip)?
Last edited by wizzy on Sat Oct 13, 2018 2:11 pm, edited 2 times in total.

User avatar
suralin
Posts: 3944
Joined: Thu Jan 25, 2018 11:36 am

Re: CS career megathread / AMA

Post by suralin » Sat Oct 13, 2018 2:01 pm

TheWalrus wrote:
Thu Oct 04, 2018 9:45 pm
Congrats dude. When do you start?

(Also, the Albert, Bernard, and Cheryl problem made me feel super dumb)
thanks! i was SUPPOSED to start 10/22

and yeahh that problem is mindblowing here's the original source http://jdh.hamkins.org/now-i-know/

Post Reply

Who is online

Users browsing this forum: No registered users and 1 guest