Hacker News new | comments | ask | show | jobs | submit login
'Clean your desk': My Amazon interview experience (shivankaul.com)
911 points by skaul on Dec 8, 2016 | hide | past | web | favorite | 509 comments

I have a confession to make. I cheat at my job. I cheat all day, every day.

I have this little book next to my desk I use to write down ideas and notes, and then I refer back to it later. Sometimes my boss is standing right there! I get such a rush.

I found this website called Stack Overflow that has so many answers to problems I run in to. Sometimes I'll just copy the code directly from the site, without typing it out again myself!

Sometimes I even just walk up to colleagues and straight up ask them for help with a problem. They just tell me things I can use in my job, out loud, in a busy office, and we still haven't been caught!

I know that my cheating gives me an unfair advantage in the job market. I know this cheating makes me an inferior programmer. And now I know I can never work at Amazon because I can't get past their super scrupulous interview process. Oh well, I guess I'll just try and get by, cheating my way though life.

Ken Robinson on collaboration in education:

"In the work world, collaboration and team work are essential to success; in school, it's called cheating."

More here https://www.youtube.com/watch?v=zDZFcDGpL4U

Well what a lot of children do in school is indeed cheating outside of collaboration. It's often one person who's already done doing the work of another without much bidirectional creative synthesis.

I would think that students who collaborate on schoolwork or studying are very institutionally fluent and aren't really at risk of the institution stepping over them and calling them cheaters.

I see the college cheating industry as a trade between wealthier cooler kids and the smarter disadvantaged kids, which I suppose is also life in most places, China or US.

Well, cheating is only to be expected in a system that puts a score/grade on the pedestal.

The point of school (American school, at least) isnt to learn, its to get a good grade. When the goal is a number, than 'cheating' is simply trying to achieve the goal.

If the goal was to actually learn, the "cheating" that you see in schools would become "collaboration" - much like in a workplace with quotas and ridiculous metrics, people will cheat the system - in a workplace who values output quality first, you end up with collaboration between employees to reach that goal...

The problem is - 90% of us went through the school indoctrination of "the grade is the goal" ... and carry a lot of that mentality out into the work force.. and at worse, many of them start companies and align their business goals for their employees with that same standard..

Though I don't disagree with your argument that grades weigh to heavily in American education, I disagree that "collaboration" and "cheating" are the same. There are certain, rote skills you need to take away from school and on which you need to be evaluated individually. You must know how to read, do arithmetic and algebra, understand civics, geography, etc. The problem with encouraging collaboration over evaluation is that evaluation is still necessary to determine whether kids are learning what they need to learn.

I think we all remember that kid in school (or college) that joined a group and coasted to grade. I know those people where I work, too. They know little and contribute nothing. We do not serve them by letting them get out of school without the knowledge they need.

So I agree that grade-centric education is poor, but can also not reject either rote learning or evaluation as being an important goal.

The point of school (American school, at least) isnt to learn, its to get a good grade. When the goal is a number, than 'cheating' is simply trying to achieve the goal.

Aka Goodhart's Law[0].

[0] https://en.m.wikipedia.org/wiki/Goodhart's_law

The education system ought be multi-purposed. Some people want school to get them a job. Some people want school to give them a 'well rounded education' on matters like history, American civics, math, sports, sexual education, etc. Some businesses want schools to correctly identify and promote talent, for better hiring signals. Some people view school as a way for preparing for college, where a student might choose specialities.

These are all goals which are harmonious to societal health.

Burning my karma, but I can't resist; lots of students in my high school agreed that collaborating in sex ed was far more beneficial than studying solo.

I don't get it

When he refers to "collaborating in sex ed", he isn't actually referring to studying, but rather to teenagers having sex. Similarly, "studying solo" refers to masturbation rather than actual studying.

Having sex is better than masturbating.

Really? He means intercourse is more fun than masturbation. HTH!

The education system boils down to attempting to certify that someone knows what they claim to know.

This so that said someone can be slotted into place in the company production line.

Yep. Education isnt the goal, menial job placement is. Its cynical, but also true, that an educated populous would grind America to a halt, because we're reliant on people doing a lot of menial jobs... and educated people dont do those jobs...

> Education isnt the goal, menial job placement is.

I disagree. Plenty of times through school I heard complaints of "When will I need this in the real world!" and numerous comments online complaining about education (particularly arts and social science curricula in university) not preparing people for jobs.

Plenty of things I learned in primary school, high school and in my Engineering degree are not applicable to my day job but did make me a more educated person.

It's funny to witness the education system get simultaneously criticised for not preparing people for the workforce and also only preparing people for the workforce.

> Plenty of things I learned in primary school, high school and in my Engineering degree are not applicable to my day job but did make me a more educated person.

It doesn't matter. There must be something that can be taught to you and later used to grade you. Whether it's maths or arts or history of the good deeds of Comrade Stalin, it doesn't matter. What matters is the series of filters, at the end of which is entering the job market. Even the parents don't care what you're being taught - only that you are good at it, so that you can go to better higher-level schools, and get a better diploma at the end. Only few parents and teachers with ideals give a shit about the actual content of the education material...

As for job usefulness, 90% of stuff you'll have to learn on the job anyway.

As for job usefulness, 90% of stuff you'll have to learn on the job anyway.

Of course. Education is about laying the foundation, giving you a bounty of contexts with which you can make faster associations and more quickly learn what you need for the job.

"Plenty of times through school I heard complaints of "When will I need this in the real world!" and numerous comments online complaining about education (particularly arts and social science curricula in university) not preparing people for jobs."

Both of these are correct - and they corroborate my point. The skills taught, they don't prepare you for a real world job - they teach you to stay in line, as few questions, and do what you're told. Thats what you need to be tought to work in a factory or plow a field... or stand behind a cash register.

They don't prepare you for "good" jobs... they prepare you for menial jobs. If everyone came out of high school with real knowledge and real skills, nobody would run the cash register at the gas station for minimum wage...

Not sure why you are being down voted. Perhaps people are just bitter at school.

What a lot of adults do at work is also indeed cheating outside of collaboration.

"In school, taking credit for someone else's work that you had nothing to do with creating is unethical, not accepted, and punished severely. In the work world, it's called Tuesday."

--Anyone who's experienced it

Cheating and collaboration are not the same. If you've ever taken an advanced math/cs/STEM class you've collaborated often on problem sets, they are often too hard to do alone. But you shouldn't have cheated.

Collaboration is great for learning something new as a group.

But how do you then assess whether each individual in the group learned that important concept, or if they just all agreed with the one (and only) student who actually learned/knew it?

It probably requires a shift in what we think of as assessment. Montessori has some ideas:


Seems to produce good results:


What would be the point of assessing an isolated individual in a world of collaboration?

Its akin to evaluating how someone would do something in the middle of the jungle without tools or support. Sure, you will be able to know who does better in that setting, but what conclusion can you bring of that?

It's gonna kinda suck to be the engineering team where nobody actually learned calculus.

Or the surgical team where everyone "collaborated" on basic anatomy.

How would you know nobody learned calculus?

You are comparing people that dont know to people that know with that allegory. In a collaborative setting you would be able to see what people build, teach each other, and learn.

A team with 5 people that know 20% calculus can produce much better results than a team where 1 person knows 100%. And sometimes they wouldnt be able to produce a result at all.

> but what conclusion can you bring of that

Welllll you could conclude that they are in a prime position to create a YouTube channel with 178 million views /s


>>What would be the point of assessing an isolated individual in a world of collaboration?

Collaboration can only occur when each party has something meaningful to contribute.

Yes, you're absolutely right, that someone's effectiveness -- in academia or otherwise -- depends on those around them.

But you still have to distinguish those who are not effective at all, contributing nothing, but who are still putting their name on the project.

To use the sibling example, there's a difference between a surgeon who draws on other specialists in the operation, vs someone that knows nothing about surgery.

> But you still have to distinguish those who are not effective at all, contributing nothing, but who are still putting their name on the project.

You dont need an individual assessment to do that. The real work world is exemplary on that. People without degrees very often outrank and outperform people with degrees, either through curiosity or through other skills that make him a more important key member.

In real life if someone is regarded as unproductive by his co-workers is not re-tested individually for validation.

Both collaborative assignments and assessments of those assignments are often poorly conceptualized in school.

Why, for example, do curriculum developers create (or choose to adopt from textbook publishers) scenarios where if one person fails, everybody in the group suffers? That's not why we collaborate. In fact, that's the opposite of why we collaborate. We don't do it to create multiple points of failure.

It's also a problem with the way we assess and the fact that we almost exclusively conduct summative assessments on these types of projects with high-stakes consequences rather than formative assessments that allow the experience to be used to foster understanding.

I'm a parent (and an education journalist). I see a lot of this. My daughter has had to do the entire work for several "collaborative" projects to maintain her GPA just because the other students wouldn't do the work or wouldn't do it well. That's a common complaint.

Two ideas for getting around that:

1. For the types of assignments where a small group produces a single outcome and shares responsibility for it, assess formatively. If the group does poorly, the teacher sits down with them and gives them supplemental instruction. And, in talking to the students, that teacher can also discover where the weaker students are and focus on them. That way you get assessment with zero high-stakes consequences, and you help advance learning rather than ending it with a letter grade.

2. Assign projects where each student has a job and is evaluated singularly for that job but is allowed to work collaboratively with other students (and the teacher) to complete it. There are all kinds of ways to encourage collaboration in those cases without resorting to punishing others for one person's failure in the end product of a collaborative assignment.

As others have said, there's also the question of assessment itself. Why assess for a grade or points or what have you? There's a movement to do away with grades and test scores. It's working fairly well, but it requires serious dedication from the teacher to ensure that students are learning. Large-scale, that's difficult. We don't have 4 million+ teachers who would be willing to do that extra work.

In University for me at least, a lot of those collaborative projects allowed the students to distribute their grade amongst participants.

In my experience at least the weaker contributors to the groups success, even those that just coasted, were still very honest when asked to assess their contribution and the contribution of others.

In my university we often had group projects and at the end of the project you scored your team members. If the entire team scored one member highly or poorly, their grade could be weighted higher or lower.

Furthermore, how do we ascertain whether or not group work is just "outsourcing" for teachers and the students that learned the concept aren't simply k-12 unpaid TAs who find their role as "assistant professor" pretty tedious thus turning off the best and brightest from a lifetime of enriching educational experiences?

Actually you can learn a lot from teaching others or helping others learn the material. When you have to teach someone else something you learn what you don't actually know that well.

I remember having a cheating rig at high school. We were 4-5 classmates, who would always sit close and hangout. Each of us would study the classes we best at. For example I really sucked at English and math, but was really good at physics and biology, one guy was good at English and another one at Turkish literature, another at history, another at Math. Overall high school was really easy for us. It's not like we didn't cheat at exams, we also helped each other before the exams. Teachers tried to separate the group several times but we were smart and doing well, so they chose to ignore at some point. All members went to uni, studied STEM.

Now, I'm well aware that studying things you can already do is not the best way but this was what happened.

Until you get slapped with a parent or copyright lawsuit...

Sir Kenneth Robinson! :-)

Similar to your little book, I have a personal wiki in a Git repository which basically functions for me as an external memory--if I find a solution to a problem that keeps coming up, or write a particularly useful snippet of code, I put it on the appropriate page so I can remember it again later. (When I say 'external memory' I really mean it--in addition to pages like Snippets:Bash I also have things like LifeSkills:SuitcasePackingList.)

Something that worries me is: what if I work for a company that decides that everything I've written on their time belongs to them? If I had all of this in my head there would be no question about it, but writing it down introduces questions of copyright. Should this ever happen it would basically be the equivalent of losing a chunk of my brain, which is a bit of an alarming prospect.

I work for a company that decides that everything I've written belongs to them. Note that I did not include the phrase "on their time." I have to get explicit waivers for any project I want to work on on my own time. It is more than a little disconcerting. I'm pretty sure they technically claim ownership of this comment.

Yes, draconian IP agreements suck!

What can you do about it (if you live outside of California)?

Your key weapon here is knowledge. Specifically, the fact that California Labor Code section 2870 exists.

"...shall not apply to an invention that the employee developed entirely on his or her own time..."

Here's what you should do to avoid draconian IP clauses:

* Print out a copy of 2870 and bring it with you when you go to sign your job offer http://webcache.googleusercontent.com/search?q=cache:3qLjxOo...

* Read your contract. If you don't see similar wording in their contract, hand them your copy of 2870 and ask it to be added to the contract.

* If they try to push back, calmly explain that every resident of California gets this "for free". If Google, Facebook, etc employees all get this, why shouldn't you?

* Be prepared to walk if they refuse.

But what if you're already in a job which has draconian IP clauses?

* Try bringing 2870 to the table the next time you are negotiating for a raise or promotion.

* If you are cheeky, go get a job offer from a competitor, and highlight the fact that they are willing to offer you 2870.

But the best thing you can do is to spread the word about 2870! If enough of us demand this, eventually it will be such a hassle (for employers outside of California) that no one will bother trying to trick their employees into draconian IP clauses.

I should give a hat-tip here to my former employer, uShip (Austin, TX). They include this clause by default for all devs, and I'll make sure I ask for 2870 at every future job.

Just because they make that claim does not make it true or even enforceable.

Generally anything you write on company time or using company equipment belongs to them though.

You must have passed the Amazon interview.

That is absurd. You should quit and find a new job. Never sign away your rights to create IP on your own time.

Honest question... how would the company define "on their time"? You are (presumably) salaried, and it's not like we work in a 9-5 industry or can't do work at home. IP is important to them, how else could they structure the arrangement? These "we own your ideas while you work for us" clauses are common.

Most contracts I see these days define it as work that could be reasonably seen as in their line of business, or using company equipment or resources (e.g. internet). I can't remember the last time I saw one that made a distinction based on clock time that wasn't also for a 1099 contractor. These days it seems it either goes "All your thoughts are belong to us" or "using our stuff in a market we sell to".

A reasonable definition of "on their time" for a salaried employee would be things they asked you to do for them. (Presumably they did not hire Arandur to write HN comments.)

Another IMO reasonable definitions:

- you used company equipment for that == it's ours

- you billed us for the time which you spent on doing that == it's ours

Did you not query that clause in the contract?

In my first developer job out of college, I queried that clause, and they basically said the job was conditional on accepting the clause.

And you took the job? Why?

In my case, these are the justifications that I came up with (in no particular order):

1. I liked the people

2. The technology was interesting

3. I felt like I needed to break into the industry as a fresh grad

4. Every other company I'd interviewed with had a similar clause

5. The side projects that I work on are strictly non-commercial, for my own education, tend to be difficult to monetize legally, and have nothing to do with anything that my employer is interested in.

6. (found out later) My employer doesn't enforce its anti-competition clauses or try to take over IP that employees develop on the side.

I haven't regretted it in 8 years, although I'd imagine that a lot of the people on HN would be a lot more entrepreneurial than me.

When looking for a job, one does not always feel in a position to turn down a job offer, especially when one does not already have a competing bid.

Can you share this repo? I really like that idea.

I don't know about his example, but I'm using gojot[1] for that. Automatic encryption, entry management, and syncing if the repo is remote.


I'd rather not share the actual repo, it's a mess and also has some semi-private stuff in it. At some point, I may go through and clean some of this up and publish it, but for the moment it really is an extension of my brain--imagine publishing a feed of your thoughts directly to the world (Twitter notwithstanding).

I can, however, give you an idea of how it works. There are many personal wiki programs out there; I settled on Zim[1] as it stores everything in plain-text files and supports a basic markup. On my phone I have the repo set up with Pocket Git[2], and I view it with Coastline[3] (which is just an arbitrary text editor; I don't mind reading the Zim syntax). The Git repository is stored on BitBucket[4] since they offer free private Git hosting, though I now have my own VPN and am considering moving the repo there. Syncing is done periodically (roughly once a month) by committing, merging, and pushing, usually with a few minor edit conflicts to the more popular pages.

Pages are organized into a very loose and fluid hierarchy; for the most part I put things someplace that seems reasonable and then periodically merge, split, and move pages when they grow too unwieldy. The major parts of the hierarchy are:

- Project:(Active/Complete/OnHold): major projects (anything from Barcamp2015Presentation[5] to NewCellphone to BetterWorld[6]) get a page, which collects a narrative about the progress of the project, a summary of the current state, and any relevant links or quotes; this page may have subpages as well.

- Entity:* collects links for anything related to someone or some company I've been involved with; mostly this is backreferences to projects I've worked on for them etc. For example, Entity:Person:JohnSmith could link to Project:Complete:JohnMidiController for a side project I did for a friend.

- Hardware:* collects everything hardware-related, for example Hardware:Computers:RaspberryPi has links to pinout diagrams and the like.

- Software:* collects software-related pages, with subsections for Software:Protocols, Software:Programs, and a smattering of loose pages.

- Snippets:* covers code snippets, such as Snippets:Bash, Snippets:Python, and so on.

This covers about 75%; the other 25% is somewhat less organized and consists of things like fiction ideas, links that don't fit anywhere else, the aforementioned LifeSkills namespace, and Miscellaneous:Quotes:FunnyPhrases ("you there, wear these groceries"[7], "Turns out I'd just existed inconveniently at you"[8]).

Hopefully this rather long post answers your questions, but let me know if there's anything more you'd like to know.

[1]: http://www.zim-wiki.org/

[2]: https://play.google.com/store/apps/details?id=com.aor.pocket...

[3]: https://play.google.com/store/apps/details?id=com.rumsunrise...

[4]: https://bitbucket.org/

[5]: http://www.linestarve.com/blog/post/A-Mad-Scientists-Guide-t...

[6]: A Grand (Semi-)Unified Plan for improving everything, encompassing BetterUserInterface to BetterGovernment to BetterEconomy.

[7]: http://www.wildelifecomic.com/comic/195#comment-2550773663

[8]: http://suburbanjungleclassic.com/?p=1207

Today I used "man" to check some parameters of a command. I felt such a rush knowing that I cheated the system.

I think I know what you mean. I had to run a few commands using the `--help` flag this morning to look up features that I wasn't sure they could do. It just feels so _dirty_.

At least we know now that we can't be counting on you in situations of dire need - like https://xkcd.com/1168/.

If you want to cheat a little more you can try tldr: https://github.com/tldr-pages/tldr/

It gives actual examples of commands, just like a cheatsheet does.

This is great, thanks for the link!

whoa, never knew this existed. Thanks for sharing!

Stick it to the, er, authorities!

> Sometimes I'll just copy the code directly from the site, without typing it out again myself!

I realize you're being facetious, but do check with your employer that this is OK! Code on StackOverflow is under a CC-SA license, it isn't public domain or even very permissibly licensed.

I was like "can't be true" but it seems you're right:


The license [1] says that you are free to:

Share — copy and redistribute the material in any medium or format

Adapt — remix, transform, and build upon the material for any purpose, even commercially.

The licensor cannot revoke these freedoms as long as you follow the license terms.

Attribution — You must give appropriate credit...

ShareAlike — If you remix, transform, or build upon the material, you must distribute your contributions under the same license as the original.

[1] https://creativecommons.org/licenses/by-sa/2.0/?

It's not entirely true:

> Starting Feb 1, 2016, all new code contributions to Stack Overflow and Stack Exchange will be covered by the MIT License.

Source: http://meta.stackexchange.com/questions/271080/the-mit-licen...

So, it depends. Code from before then is CC-SA, after it is MIT.

That post has an update at the top that the license change was delayed until March 1, 2016 and a link to a follow-up post: http://meta.stackexchange.com/questions/272956/a-new-code-li...

That follow-up post says that they're "going to delay the implementation", so I assume that today it's still all CC-SA. I haven't found any further information, but would love to know it, or be corrected on my assumption.

My company is struggling with the code license situation and I'm stuck with dealing with it for my project. Clear, up-to-date information about it would be fantastic.

Oh nice catch, I missed that part. And I feel your pain. Starting a similar project here to inventory 3rd party code etc., it's going to be hard.

But usually, code on SO is kinda like the docs, they show you how to do something and is often more of a template than a complete solution (unless you asked the question). Its not like people are writing complex solutions and putting them on SO. I mean how are you going to copyright this? (taken from a SO post):

for (Map.Entry<String, Object> entry : map.entrySet()) { String key = entry.getKey(); Object value = entry.getValue(); }

Or, famously (in a case where people thought carefully about both copyright and patents):

  private static void rangeCheck(int arrayLen, int fromIndex, int toIndex) {
    if (fromIndex > toIndex)
	    throw new IllegalArgumentException("fromIndex(" + fromIndex +
            ") > toIndex(" + toIndex+")");
    if (fromIndex < 0)
	    throw new ArrayIndexOutOfBoundsException(fromIndex);
    if (toIndex > arrayLen)
        throw new ArrayIndexOutOfBoundsException(toIndex);

This is one of the reasons I followed the Oracle v Google suit with such interest.

Below some minimum threshold of complexity, it seems like software ought to cross the idea/expression divide and become copyright-exempt, just as a mathematical formula does. In fact, the example you gave is something that could easily be recreated letter for letter without seeing the original post.

I'd be utterly shocked if that sort of "could have written it from scratch" code ever caused someone an IP problem.

>"I found this website called Stack Overflow that has so many answers to problems I run in to. Sometimes I'll just copy the code directly from the site, without typing it out again myself!

Sometimes I even just walk up to colleagues and straight up ask them for help with a problem. They just tell me things I can use in my job, out loud, in a busy office, and we still haven't been caught!"

I've lost count of the number of (thankfully former) co-workers whose only technical skills were searching Stack Overflow and asking their colleagues for help with a problem. The measure of a programmer is what they are able to do when there _is_ no answer to be had from SO or their colleagues.

I don't think there is an accurate measure of programming skill. Experience, diversity, challenges faced, skillfully using solutions to past problems, effective communication, reaching out for help when needed (including StackOverflow) are qualities that I can think of from a programmer that I would pay to build my product. I work / have worked with co-workers that you are trying to point out but use of StackOverflow was not my measure.

Something not being on SO is normally my first indicator that I've stumbled upon something interesting. I consider it a good thing.

I hope you contribute the questions and answer to SO then! Rising tide lifts all boats!

No, the measure of a programmer is the change in productivity a company suffers were the programmer to leave.

Some programmers work on bespoke algorithms and would do well to listen to the relevant commentators in regards to best practice. Not to mention that trusting everything you are taught can severely stunt your ability - read people's experiences and results.

I get that you're being facetious, but just to take you seriously for a moment: if the help that you're asking your coworkers for is something like "tell me again how pointers work in a linked list", there's a problem.

Why? Maybe they didn't understand it when that coworker explained it the first time around.

...Because if the coworker had to explain it the first time, you probably aren't qualified. Depending on the job, of course.

Hey, don't let too many people know this, but nowadays you often don't even need to copy/paste the code! For example, I have to parse Excel files a lot in my work. Did you know some dummies have already written and tested a bunch of code that does exactly that? And then they place it online in a format that's easy for me to download!

I really hope they don't realize that I'm using their code to cheat.

Reminds me of this post a while back... https://news.ycombinator.com/item?id=11700640

"Am I really a developer or just a good Googler?"

We specifically have an interview question here which is: "How would you feel about asking for help, either from a colleague, friend or on the internet, to complete a work project?"

(pro tip; the correct answer is, "fine")

Welcome to the club! No shame to be had here. Freedom and responsibility.

> I found this website called Stack Overflow

You missed the part where a search engine shows you several links, including the ones from Stack Overflow.

GP might be cheating, but not that hard.

> shows you several links, including the ones from Stack Overflow.

And all of the stack overflow answers that were linked from the top of the google search are "Have you ever heard of Google, N00B?"

I'm also a big time cheater. :)

Amazon SDE here. The SDES internally are PISSED about all of this, and I assure you many people are escalating with HR to have this new ProctorU-based interviewing process changed ASAP.

edit: I don't know if there'll be an official announcement, but as of right now we're pulling usage of ProctorU for intern loops.

For those asking how this happened, you simply do not understand the THOUSANDS of interns Amazon needs to interview every year over a couple of week period. It's a nightmare to scale. So, someone in HR thought they'd show some bias for action. Oops.

I'm glad you're fixing it, but having been through a second round interview where I was flown out to Amazon in Seattle. I will still never interview with Amazon again. It was the second most frustrating interview I've ever been in. It was clear to me that 4 of my interviewers had no intention of even considering me.

Side note: The most frustrating interview I've ever had was with Microsoft Boulder: they just flat out insulted me in the interview and questioned why anyone would hire me based on not getting their trick question. That said my interviews with Microsoft Redmond were lovely.

I went for an interview in Redmond, for which the recruiting agency had clearly 'oversold' the job description (I should have known better, too) - it was billed as more of a PM position than what it was - line / UAT testing.

Anyway, I got into the interviews, and I could sense something was up. I was answering the questions but they didn't seem... enthusiastic... about them, or me. I like to think I usually interview decently, though of course I can improve.

Eventually someone says, "Hmm, can you wait here a minute?" and a new person comes in with him a few moments later.

"So... I'm not sure why the agency sent you to us." Okay...? "You're definitely over-qualified for this role, and frankly we think you'd be bored." I thought it was a good opportunity to get in the door (and MSFT is a great place to work), so I tried to offer a little placation, when he introduced the other person, "But I know that [name] here has been looking for a PM, so why don't you talk to her."

I walked out that afternoon with a different, better, higher paying job, because someone thought "outside the box".

> because someone thought "outside the box".

That's "thinking for the company" which I had rarely seen during my 10 years working for a large company. "Thinking for myself / department" approach is much more common, unfortunately.

True, true, too.

But I just appreciated that could have so easily been a 'thanks for your time', handshake and leave, with no-one satisfied.

I'm suspicious based on something you said about the interviewers being disinterested and unenthusiastic. I am cynical, having seen this process played out many times, and having wasted more than a few days interviewing for jobs I had zero chance of earning because the employer had already (yet unofficially and potentially illegally) chosen a candidate. Let me explain:

What you've described sounds to me like a setup for hiring an H1B employee. The employer is legally required to interview everyone who meets the requirements, and must hire qualified domestic employees if they are available for work, but the employer will look for any reason to disqualify domestic candidates.

A whole industry of consultants popped up a decade ago that coached companies on how to use this "loophole" to favor H1B over domestic employees (which is on it's face, illegal. If a US employee is available, the employer must hire him so the trick is to find a way to disqualify this domestic candidate.)

Or maybe they already had an internal candidate and were required by HR to interview some additional people (e.g. you) before they discover, to everyone's surprise, the internal candidate is the very best one for the job. "Going through the motions" to interview others when their mind was made up before the job description was even written.

EDIT: Just so it's clear, I'm not saying Amazon does these things; and I don't know anything about this particular case. I am saying this sort of faux-interview happens in the industry, and has the same "feel" of the interviewers simply going through the motions.

This is a common misunderstanding about H-1B. For hiring someone on H-1B, the employer need not prove that they made attempts to hire a citizen or permanent resident first. But to sponsor an employee on H-1B for green card, they have to provide such a proof. What you describe happens during that process - you have an employee you are happy with, want to sponsor her for permanent residency in the country, and are willing to go through the expenses, but state insists that you have to try find another person first.

The right solution for this is to keep employers out of the skill-based immigration process.

That's a perfectly acceptable concern to have. Especially with Microsoft, and its use of the big three agencies, H1B in general, perma-temps, etc.

However, there was a bit more to the story - one, that I'd started discussions with them and the agency prior to moving from Australia to Seattle, and they'd been willing to sponsor a visa (that wasn't needed), and sure enough, a couple of weeks after, my team had some new testers who weren't H1B who'd filled the position.

And it might be a little misleading on my part, there wasn't a sense of absolute, outright apathy. But just a growing more deflated tiredness, perhaps.

That's a funny story, thanks for sharing.

Can you share some details about why you found it so frustrating? Believe it or not, most software devs at Amazon REALLY CARE about interviewing, are happy here, and want to find others too.

We are taught in an internal interviewing class (which is not mandatory unfortunately) that making sure the candidate has a great experience is just as important as getting good data on the candidate.

>We are taught in an internal interviewing class (which is not mandatory unfortunately) that making sure the candidate has a great experience is just as important as getting good data on the candidate.

That's funny. I don't know a single person who enjoyed interviewing at Amazon, and the one guy that did who got in quit a few weeks later and still won't stop talking about how bad it was until he's blue in the face. It was such a bad experience it seems to have partially fused with his identity!

Seems like every once in awhile there's some article taking a crap on Amazon, and then these types of green comments coming to the rescue.

Ex-Amazonian here. Amazon is big enough that there are actually a lot of people in the org who enjoy their jobs, think the company's generally doing the right thing (albeit possibly with caveats), and don't mind defending it. The reason for the throwaway accounts is that the social media policy asks employees not to wade into discussions of Amazon, but rather to forward them to the PR team.

Addressing your experience directly - I really enjoyed my Amazon interview in Edinburgh eight years ago. Lots of interesting free-form problem solving, with the whiteboard there to jot down notes rather than to produce working code.

I'm guessing there are teams or entire orgs which just don't take interviewing as seriously, though - again, with Amazon's scale, it's not surprising.

I don't doubt what you say is true, but Amazon is the only big-3-ish company that I regularly hear some story like this about.

I think every company has a mix of horribleness and goodness, and based on the data I have, it seems like the mix at Amazon is more skewed towards horrible than other major companies.

(disclaimer: Amazon employee)

I've made the same observation as you, but I think there are other factors at play. It seems that Amazon has become the new "cool" company to hate, just like Microsoft was a few years ago. I've personally heard a mix of positive and negative stories from all the big 3, but Amazon seems to be the only one that makes headlines.

I'm not trying to say that Amazon is perfect, or even better than Microsoft or Google, I'm just saying that maybe it is better to trust the opinions you hear from people you know rather than what makes the front page of HN, wherever that happens to take you.

During my Microsoft internship, I watched a male coworker verbally harass a female coworker to tears while she was giving a presentation and no one batted an eye. A friend told me to come check out Amazon, and I've never looked back. I would never go back to Microsoft, but I don't fault people for going there because I know that it is a huge company and that my experience isn't representative of the whole.

Which other big company have you heard of using interview technique described in the OP? The fact that Amazon thought up the process as described here says something about the as a company which you can't explain away as the "new cool company to hate".

Many are doing it - see the rest of this thread.

Anyways the difference here is HR/recruitment was too "ambitious".

The SDEs have found out and we have put a stop to it.

Glad to hear that the practice has been stopped. The fact that SDEs can bring up such bad PR to the attention of management and get it rectified gives one good impression about Amazon.

My wife recently went through the hiring process at Amazon. She completed an online programing test, then went through an interview where a recruiter asked basic CS questions over the phone, and was then invited for an onsite interview. She took a day off from work, and spent about 4 hours doing the interview at Amazon's SF office. The interviewers had come from Seattle. This happened over 3 months ago, but she has not heard anything from her Amazon recruiter about the interview outcome. The recruiter had earlier promised to get back no later than 2 days after the onsite. My wife's emails to the recruiter were unanswered.

We understand why things are this way - the recruiter wants to utilize her time on candidates who convert; any time spent on a rejected candidate is time "wasted". But when a candidate takes a day off from their current place of work to attend an onsite to get treated without common courtesy is sad state of affairs. Imagine if the situation was reversed - once the interview was arranged, and my wife agreed to do it, how would all those interviewers who flew in from Seattle have felt had she not turned up?

You seem to be interested in fixing things at Amazon. I mentioned this incident to point out that recruiting at Amazon can do with much course correction. Take time to look in the mirror and understand why it is that only Amazon among large tech employers attracts such negative attention. Sweeping all of that under the "we are now the cool company to hate" category is dishonest. Again, I am glad that this particular incident posted by the OP has caused a positive change in the recruiting practices.

Glad to hear it!

Microsoft's hyper-aggressive culture has contributed to its organizational dysfunction IMO, and frankly I am still having to recognize and deal with its effects on my interpersonal relations.

>>>The reason for the throwaway accounts is that the social media policy asks employees not to wade into discussions of Amazon, but rather to forward them to the PR team.

Presumably so that said PR team can create throw-away accounts and defend Amazon.

Well, possibly. I can't prove that that's not the case :) All I can say is that I've always taken them in good faith, because they match up with the temperament of some of the people I know and like inside the company.

That's really the problem with the social media policy. On the one hand, it protects the company from the worst possible outcomes of letting untrained employees loose on social media - and given the size of the company and the diverse views held by its employees, I could imagine those outcomes being pretty dire! But on the other hand there's a chilling effect on positive impressions of the company.

I'd think they should have a few non-green accounts lying (no pun intended[0]) around by now at Amazon HR PR.

[0] that was a lye[1]

[1] removing the stain on my reputation

> The reason for the throwaway accounts is that the social media policy asks employees not to wade into discussions of Amazon, but rather to forward them to the PR team.

So, they are wading into discussions of Amazon, just in a way that's not directly linked with an existing nym? Following neither the letter nor the spirit of the rule?

Hi. Not a green comment, though admittedly I'm not using my real name here.

I worked at Amazon for many years and enjoyed it. Nice to meet you.

It's a huge company with a wide variety of teams and experiences. It's not for everyone. But I am still friends with lots of my former co-workers. Lots of them liked it and, believe it or not, many of them are still there. Some for over 10 years.

I've only heard bad things as well. I respect them as a company but it's clear they don't care about employee wellbeing. Everyone I know thats worked there said the hours and deadlines were unreal.

I interviewed at Amazon and it was fine. The questions were interesting (albeit maybe a few too many graph questions). One of the interviewers showed up in a bathrobe or something I think, due to a bad oncall the previous night or something.

Given that Amazon interviews by team, I think it depends on what team you end up interviewing with...

I didn't take the offer because it was like $20k or more less than another offer, but the process was fine.

There's also going to be selection bias, in that most people who have reasonable interview process aren't going to write blog posts about it...

Not mandatory? In my time at Lab126 (a couple of years ago now), you had to take Making Great Hiring Decisions before you were ever included on an interview loop.

Some orgs make it mandatory. Some don't. There's no global policy

Huh. I had no idea. When I was there, it was mandatory in each org I was in. I had just assumed it was a company-wide policy.

I just had to chuckle at the name of the interview prep course name, of course it would be that.

This was in 2013 so it's out of date now, but it was very clear from the line of questions and the attitudes of the interviewers that they were being made to be there and saw no point. They were in general completely disengaged with me. Of the interviewers all but the last were fairly professional. The last pair I had were more junior and made it clear that I was wasting their time as far as they were concerned.

To be fair I interview poorly, but I would would have much preferred for the person managing the interviews to cut off the process as soon as they were sure that I wasn't a fit. By failing to do so it just created an awkward and frustrating situation.

Sorry to be a broken record, but can you give some specific examples?

I ask because elsewhere on this thread you have people bitching that they were asked some low level CS questions about inverting binary trees and the like, and how insulting they found that.

While I sympathize, it's not going to change the fact that the company does want to hire engineers that are equally competent at contributing to some sexy new high-scale AWS services as some boring business reporting features. Google, Facebook, and Microsoft all do exactly the same thing.

On the other hand, if you had people that were actively rude, did not give you hints when you got stuck, and made you feel unwelcome, that would definitely warrant feedback. Though, as you said, it's too late since it's 2013.

I advise anyone in the future who runs into this kind of situation to contact your recruiter with feedback, and be very clear if you had a bad experience. Amazon takes customer feedback super seriously, and when we interview, the candidates are customers.

I wish I could, but it's been three years so I don't recall specifics.

I do recall the last pair asked me to re-implement a java standard library functionality that was rather complex. It seemed out of place for an interview question.

While there are valid reasons to ask such questions they need to be framed carefully as to indicate why the developer would want to do so. Failing to do that causes frustration because in general the first rule is don't re-implement the standard library.

EX of good question: Facebook has their own COW C++ string with small string optimization because libstdc++ didn't used to have that. It can also provide massive speed improvements on multi-threaded code, what are some ways to implement this?

Ex of bad question: Implement a stringbuilder.

>While there are valid reasons to ask such questions they need to be framed carefully as to indicate why the developer would want to do so. Failing to do that causes frustration because in general the first rule is don't re-implement the standard library.

Seriously, you're bothered because someone asked you to solve a toy problem? Most interview problems are things one wouldn't do in the real job; that's because all real job problems take more than an hour to solve.

Implementing a stringbuilder is a bad question because there's not much depth to it (after you get the basic implementation down, where do you take the question next?), not because there's a stringbuilder in the stdlib.

> I do recall the last pair asked me to re-implement a java standard library functionality that was rather complex. It seemed out of place for an interview question.

Allow me to respectfully disagree.

Even if they ask you something you or they don't think you can do, it's not just about really solving it but at looking at how you approach such a problem.

No one can really solve important, long term problems in an interview but you sure can reason about them and talk/try toy solutions.

> Ex of bad question: Implement a stringbuilder.

You can always say "I probably wouldn't do this unless I had a good reason but, If i had to roll my own, I'd approach it this way:"

We'll have to agree to disagree. I think implement a string builder is a GREAT interview question.

The point is to have a discussion. Why? What are the benefits of a string builder? When is it necessary? What performance characteristics should it have. What kind of questions do you ask, what kind of edge cases do you consider?

It's an interview - if you are not sure why the interviewer is asking you to re implement a standard library, ASK.

This is not unusual for Amazon - most tech companies these days ask these kinds of questions

> Ex of bad question: Implement a stringbuilder.

We used to ask people to implement standard date-related functions. "That's stupid, all the date functions are already written, and they handle all the weird edge cases for you!" That's right - we want to see if you can think about all the weird edge cases.

It's not about the obvious functions, it's about whether or not you can plan reasonable interfaces and think about edge cases.

How about escalating to Bezos?


I was going to interview for amazon. The recruiters were so disorganised, I just gave up.

> We are taught in an internal interviewing class (which is not mandatory unfortunately) that making sure the candidate has a great experience is just as important as getting good data on the candidate.

Was that not in the mandatory one?

I'm curious, what was the trick question given to you by Microsoft Boulder?

I don't recall the exact question, but I recall they were only looking for the answer that used bit manipulation. Which was not necessarily intuitive in the context.

I know a common one is how to switch the values of two variables without using a temporary third variable. One way (method 2 here [1]) is to use repeated bitwise XORs.

Does that sound like it could be the one?

[1] http://www.geeksforgeeks.org/swap-two-numbers-without-using-...

Hm, perhaps counting number of bits set to 1 in a multi-byte array?

no, bit wise is logical there. I'm thinking it was duplicates or primes or some such.

Glad that SDES internally are upset but the damage has already been done. Amazon is becoming a place where serious tech talent will not work. No senior or serious talent in their right mind would go through this, so you are going to end up with just junior or desperate devs applying.

Not to mention the turn over. What is that like in most tech depts?

I suppose however, Amazon can simply use the excuse of: we can't find qualified devs so we need more H-1B workers. Heck I am starting to think that this is the plan all along.

I apologize if I am coming off as negative but this kind of stuff really has to stop. I don't understand how it got to this point and how anyone would be willing to give up their privacy or self-respect like this. How did we get here?

> Amazon is becoming a place where serious tech talent will not work.

It's not becoming it. It already is.

Doesn't everyone have conversations with their tech friends about how everyone who's been there is trying to go away and interviewing at Amazon was terrible?

I'm not even kidding. That "never work for amazon" is the image we have from amazon, it's to the point it could be a meme.

I really home recruiting at Amazon takes this comment to heart. They won't, because they're too busy milking the college newgrad pipieline and that looks fat and happy, but experienced engineers in general won't work or even interview for Amazon. You get the naive people, everyone who knows better stays far away.

> Amazon is becoming

It very much already is a place no serious tech talent will work. At my last place of employment, we used to crack jokes about working at Amazon.

Someone was probably promoted for the adoption of this "frugal" hiring practice. This is a culture problem that permeates all of Amazon and goes all the way to the top.

Find a new job. The reputation of your company is going down the drain.

With all due respect, you don't know what you're talking about :)

I, and most people I know have mostly positive Amazon experiences. We are proud of the work we do, and the innovations of the company. I interview a LOT, and I make sure that 100% of my candidates come out with a positive experience, even if they do not get hired.

There are problems at every company, but right now it's fashionable to shit on Amazon, ever since the NYT hit-piece that got a TON of facts straight up wrong, and ever since then every negative Amazon story gets upvoted to the stratosphere. Every company makes mistakes, it's how you deal with them that counts.

Oh and in this case it wasn't a question of frugality, but scaling. Too many interns to interview, not enough time. It was still a bad decision, but it wasn't about being cheap.

Listen, your hiring process reflects your company culture. The fact that HR is in charge and you have to work to change it, rather then you, the people who will be teaching and supervising these interns, sitting down, telling HR "This is what we need, This is what we want, how can we make this scale" and having final approval says a lot.

Further, this expectation of a private quiet clean place to interview for several hours seems rather biased against poor people. Not to mention the test itself sounds like it doesn't account for the possibility of people with motor disabilities, the blind, the deaf and other disabled people.

Honestly, open book tests are a thing that has been extensively studied. Use IRT based adaptive multiple choice tests (like the computerized SAT or GRE where tests adapt to your ability), set a time limit, have a rolling set of questions to prevent knowledge transfer, and have an intern test day every quarter. You can calibrate test responses based on Amazon engineers at their desks.

> rolling set of questions

This is the hard part, and I think it's why open book tests don't happen more. Coming up with good questions is hard. Don't forget that if they're not sufficiently work-like people on HN will still say that your hiring culture sucks.

I haven't worked at Amazon, and I won't speak to whether things are uncommonly bad there. Certainly Microsoft seems to have produced plenty of similar stories while escaping the broad stigma Amazon has acquired.

I will say, though, that I think your bad-reputation timeline is totally off. By the time that NYT piece broke, I had heard "Amazon is a horrible employer" from a half-dozen different places. I didn't even finish the thing, I just shrugged and went "well, matches what people have told me, no new info here".

The first time I was warned away from Amazon, it was over specific (and probably unenforceable) pieces of their internship contract. A friend rejected his offer because it contained a bizarrely broad non-compete that could theoretically have barred him from ever starting a company with anyone else who had ever worked at Amazon. This was 2012.

The second time was from a friend-of-a-friend who spent one year as an SDE. He talked about bureaucracy, burnout, and made the Microsoft comparison. He said it Amazon was reasonable as a ~2 year cash out, but nothing more. This was 2013.

After that, I was still open enough to respond when I heard from a recruiter. The internship interview required invasive rules like the ones at issue here, which proved fundamentally incompatible with my machine/internet setup at the time. I declined it, and scratched off Amazon unless I got news that these things had improved. This was 2014.

How you deal with mistakes is what counts, agreed, but Amazon's reputation struggle is far older and more pervasive than the perception spread by the NYT story. Fair or not (and I honestly don't know), it's a long-standing issue.

I interned at Amazon in 2014 and in 2015. I have not worked there since.

The non-compete and non-solicit for customers & business partners lasted for 9 months from when my internship ended and the non-solicit (titled "Non-Interference") for Amazon employees, contractors & consultants lasted 6 months from when my internship ended.

The interview process was two back-to-back phone screens. I heard from other interns the following summer that it had changed to an online coding test and one phone screen. I had never heard of anything so invasive as what is mentioned here before this article (and the precursor a week or two ago) hit HN.

I was working at Amazon when the NYT hit piece broke. It read like it was about the business/sales/marketing side of the company and it did not at all reflect what I experienced at Amazon.

I have met a few interns who had bad experiences at Amazon but most had good experiences and were invited back. I can only think of one who was invited back but had such a bad experience that they refuse to ever work for Amazon again.

When it comes to things like this, the variation between teams within large companies is much much greater than the variation between large companies.

I agree. If you read the glassdoor reviews, engineers were posting warnings about Amazon back in 2008.

I don't work for Amazon or have any affiliation with the company, but I can confirm that everyone I know who does work there (or used to) thinks highly of the company overall, and more or less enjoyed their time there. This throwaway account seems to be fairly balanced in perspective.

I'm personally of the opinion that Amazon is not a perfect company, but reports of its "evilness" have been greatly exaggerated. Moreover, I think this is a case of people with negative experiences having more of a reason to chime in than people with positive experiences. That somewhat biases threads, unfortunately.

I'm throwing this comment in here because I think it's easy for people to forget than n=1 anecdotes don't really confirm or deny anything (positive or negative) about a company's overall culture.

I've been told opposite stores from my n=2 sample size. One has said it's miserable and he couldn't wait to get out, the other that it's great and his group is doing really interesting work and I'd love working with him.

All of your arguments would apply to any other big tech company like Google/FB/MS. Then why don't we hear as many negative stories about them?

Amazon can be worse than GooFaceSoft, thus getting more stories, and still be pretty good.

My impression has been that Amazon is highly variable internally, a bit like MS (which I actually do read horror stories about).

As someone who left Microsoft for Amazon, most of the things I hear about Amazon align much more closely with my experiences at Microsoft. We're talking about companies that employ thousands of people, you're going to hear different stories from different parts of the company.

It's entirely possible for a "publicity bubble" to cause all this. A big article got published ripping Amazon, got a lot of shares and views. For a while after that, anything negative about Amazon that seems authentic gets upvoted hard because it's what people want to read, and it lets them feel what they would call justified rage. People like that and people who just want forum karma and attention aggressively seek out or sometimes even make up stories to meet what these upvoters want to read. Stories that buck the narrative get downvoted and ignored. And just like that, you form a big popular impression of something that may be false or grossly exaggerated.

I don't know that the impression is actually false in this particular case, but these things can happen, and you'd be wise to not take the "internet consensus" too seriously.

> It's entirely possible for a "publicity bubble" to cause all this.

Possible, but also staggeringly unlikely. Occam's Razor suggests it's a shit place to interview, nothing more complicated necessary.

> Oh and in this case it wasn't a question of frugality, but scaling. Too many interns to interview, not enough time.

I've interviewed for Google and Apple internships and the process has been extremely pleasant, with the interviewers happy to give their time. With Apple, I got to meet the entire team and spend time with them. I've heard similar stories about the Microsoft interview process (I mean, they even fly you out to Redmond).

What makes Amazon incapable of interviewing their interns adequately?

The only people who I've met in real life who say that Amazon is a good place to work are people trying to hire me (with one exception). Everyone else tells me that either that they hated it and quit, or that they hate it and will quit soon. I work for a competitor so there's absolutely some bias in the people I'm exposed to, but the complete lack of positive experiences related to me by humans in real life is surprising and compelling.

The one exception is a friend who started working at Amazon a few months ago. She loves it. But she's also been handed a pile of money and told to basically do what she wants. They hired her to create a new product line (not tech related) and the execs apparently don't know enough to get involved so she's basically got signoff to just do what she thinks is appropriate. So she loves it. Her husband also works for Amazon and hates it.

I'm sure there are other people at Amazon who like their jobs. I just haven't met any and I'm not sure they constitute a majority.

I loved working at Amazon so much that I'm going back to work there. I've also worked for Twitter and Uber to give you a comparison.

Happy to chat about it more, this isn't a throw away account.

I had 2 friends working at Amazon and both of them were reasonably satisfied with their jobs. Overworked sure, but not depressed. I did ask them about the negative stories about Amazon's working conditions but both said that it didn't match with their experience. I know I shouldn't generalize from n=2 but I don't know how to reconcile their evidence with what I read online.

Amazon's social media policy discourages employees from posting about their experiences in reply to negative PR. I'm guessing that means that the employees who hate their job anyway disregard the policy, while the ones who care either don't take action or post on anonymous accounts :P

I was referring to private face to face conversations, not through social media.

It's a big company. How you experience varies based on what stage your product is at (greenfield? possibly being cancelled next month? Launching? Stable?), whether your direct manager is a nice person, what career stage your manager is at, and what five people you spend most of your time with.

I'm sure it's not the hell that you read about online. If Amazon were as bad as some stories indicate, they wouldn't be able to retain anyone decent. Still, I have heard almost exclusively negative stories about Amazon, even in person.

> Oh and in this case it wasn't a question of frugality, but scaling.

Just throw 95% of all applications into the bin before even reading them. That scales as good and you don't look so awful.

> ever since the NYT hit-piece that got a TON of facts straight up wrong

The one about the Seattle culture or the warehouse side? I remember both, and I remember the warehouse one being a lot of garbage too. I was working as a contractor doing Amcare stuff and found a lot of issues with the article.

And an engineer jumping out of the 12th floor because he couldn't handle the culture? You don't hear this coming out of APPL, MSFT, GOOG, FB e.t.c

Someone shot himself in Apple's Cupertino HQ. That does't mean much, as we are not even in the plural of anecdote yet, let alone data.

With all due respect to you too, didn't one of your employee's commit suicide? I think something must be wrong if so many people feel perpetually miserable there.

Attempted, he jumped and landed on a balcony 2 stories down or some such.

I don't hold that against amazonm. Any company of a size like amazon is going to have people that may be a bit unstable that get pushed over the edge.

I do however hold their treatment of their warehouse employees against them. When I look at a company I look at all their employees and see how they are treated. When a whole class of employees are treated as replaceable automation its probably the case that they will treat ALL their employees in that way eventually.

I'm glad you guys are pissed, it's insane! I just hope you have enough influence to actually change something about it; as it stands this process is horrifying.

I get contacted by Amazon recruiters every couple months or so. They can be a bit aggressive and will sometimes try to get me to attend one of their hiring events in my city. I would possibly consider a SDE opportunity but they want me to relocate to Seattle. I just feel that if you are going to spend all this time and resources trying to poach devs from other cities, you'd at least open a local office or let them work remotely.

Does MSFT/GOOG/AAPL/ETC do this or is it Amazon-specific?

I can't speak for the others, but I'll get contacted by a Microsoft hiring manager (not a recruiter) about once a year. Microsoft's recruiting process seems very decentralized; one group may have a totally different culture than another group. Some allow you to work remote, others don't. And they have offices spread around Seattle. (disclaimer: I used to work there as a consultant about 10 years ago)

Amazon is the total opposite. Like this dev said, HR has totally taken over the process. To work there you need to go through their meat grinder.

I would never consider Amazon based on what has been posted in the past, but I would also walk away from any interview process that began like this. Stupid.

The question is how it got to this point in the first place. I think it says a lot about the organization and how it thinks of its employees.

Lets at least pretend to apply Occam's Razor to things first before jumping on a company.

It's much more likely those responsible for choosing the ProctorU solution didn't realise everything around what the approach entailed, or necessarily the ramifications. That's especially true if it came out of HR/Recruiting who aren't necessarily the most technical people.

They have a very real problem, with thousands of people in the recruitment pipeline (especially during intern interview season), and they saw a way to significantly filter them out, without having to turn half the business over to more of a recruiting pipeline than it already is.

> It's much more likely those responsible for choosing the ProctorU solution didn't realise everything around what the approach entailed, or necessarily the ramifications. That's especially true if it came out of HR/Recruiting who aren't necessarily the most technical people.

OK, sure. My reasoning there would be that ultimately they don't really care. All due respect to Friar Ockham.

For some reason (in my experience) HR has a lot of power in companies. They can institute painful processes that nobody likes even director/executives and nobody seems to question it or if they do they are powerless to change it. The company I currently work for I wouldn't have a hope in hell of passing through all the hoops the interview process inflicts.

So, someone in HR thought they'd show some bias for action.

The thing is, one suspects it wasn't "someone in HR" who thought up the idea of using draconian new system that's pissing everyone off, and driving candidates away in droves. But rather, Bezos himself, personally. In fact it seems to represent the very essence of his "everything will be automated" philosophy, to the letter.

> So, someone in HR thought [...]

reads like the first sentence in most horror stories programmers go through

So was this an interview for an Intern position? Is this ProtorU thing used for anything else? Should anyone be worried about an interview for a full time non intern position? There seems to be some confusion about this in the comments.

Yes, ProctorU is also used for full-time interviews.

Source: a friend applied, and they made her use it - except it didn't work, so she ended up not even having a second round screen before getting her onsite invitation.

Blocking ProctorU shortens the process? ;)

Figure out what ports they work on and block them at the router? Sounds like a good idea...

My interview was for a full time SDE role.

OK, so then I guess my question for amzn-throw is, how does removing it from the 'intern loops' help when this wasn't an intern interview?

At the risk of making a redundant post, this is surely one of the most pertinent questions in the thread.

I appreciate the problem. It's not easy.

That said, I have to wonder whether the interviews are really necessary in most cases. For institutions that it is familiar with, I suspect Amazon could do just as well hiring interns purely based on their grades in relevant courses. If you reliably take 100 interns from Berkeley or Waterloo, say, it shouldn't be hard to figure out which courses count.

At worst, doing things this way would be much easier. Who knows, maybe it would work even better, by eliminating unconscious bias or something.

By all means keep the interview path for special cases with odd backgrounds.

I agree, fuck everyone who doesn't go to a top school.

Read that last line again.

There have to be automated ways of scaling intern hiring that do not involve this ProctorU-based approach.

If you have thousands of data points, what makes a successful candidate, and possibly, build a model around that?

But will HR change anything about this process due to this escalation?

Good luck.

It's pretty much standing policy that to be hired at Amazon you must be able to write correct code with no aid whatsoever, no documentation, not even a compiler to check your work. If they don't use proctors, they will go back to what they were doing before: having people write code on a piece of paper and then read it back to them over the phone.

or they could use the online editors like they were for years.

Thank you.

All these show what kind of candidates Amazon is looking for. A bunch of desperate people who'd do anything for [money|Amazon brand value|Whatever else Amazon has]. I wouldn't want to work at a place that resembles an irrational hell with such people as coworkers anyway. If someone is really that desperate for a job && believe they are smart to work on things Amazon scale, why wouldn't they find other jobs or heck start a company themselves solving a genuine problem? I do know starting a company is not for everyone to be able to do, but other jobs?

As someone who works as an SDE at Amazon, I find comments like this pretty funny. My favorite part of the job, and the reason I stick around, is the people I work with. Amazon AWS internally is a pretty minimal-bullshit work environment. "Irrational hell" is certainly not how I'd describe it. There are a lot of frustrations working with under-maintained/legacy systems, but that's fine.

Don't get me wrong, I have complaints, but at the end of the day I show up to work to solve hard problems with smart people, and that's very satisfying. I build tools other people build businesses on, and that's a really good feeling. I get paid what I consider to be too much, but that's what it took to get me to relocate from the Midwest (and no one I work with takes the current demand for devs for granted).

And yes, I appreciate the money, the pedigree, and the other opportunities it has opened up for me.

The reality is at BigCo there's always that one part or another part of the company that's fine to work at. There are a whole lot of other parts that are not. The trick is to find out what the good parts are and avoid the bad ones.

Maybe you just got lucky.

I 100% agree with you. It's a gigantic company. As far as I can tell (finger to the wind), my team is in the 90th+ percentile in terms of experience and maturity. I feel very lucky to work with the people I do, I've learned a lot from them.

When we start talking about the distribution of experiences of hundreds of thousands of people over a decade, we should expect a very long positive and negative tail. I expect that if I were rehired, my experience would regress to a mean worse than what I currently have.

Because not everyone [lives in the US | has a North American passport | can afford to not care about money].

I fit the exact combination you were excluding. [lives in the US with a work VISA| does not have an American passport | can NOT afford to not care about money]

I'd much rather apply to the plenty other companies that has opportunities. Seriously, IMO, this country has so much opportunities for the person that is working hard and willing to try.

Work is a place we spend the majority of our living life in the most active age of our life. It has to be at the least "NOT hell". Don't you think?

This is exactly why Amazon's behavior are so unacceptable. They abuse the fact that most people do not have those luxuries to create an unhealthy and exploitative work environment.

Hell to companies who treat their employees like assets!

To this day, i thank my dad who told me to go study in Canada rather in US since he knew how crazy immigration laws in states are. Graduating from a Canadian uni gives me options to work for any company/start up i like without having visa restrictions.

Good for you. Your dad is wise. I have a few friends who are in/approaching/just past the Green Card process, and hearing their stories makes me feel shame. These are people who have been here for years after getting their master's degrees. It's incredible.

Just curious. Can someone who graduates from Canadian university work for any startup/company in the US just like that? Never heard of this before..

Nope not in US, but i can work anywhere in Canada for 3 years which is enough time for me to get my PR.

Same in the US with post-completion OPT letting you stay and work freely for three years (including the STEM extension)

Well, 3 years (at best) to find an employer who will sponsor you for an H-1B visa application. And then you have win the once-a-year lottery to get the actual H-1B visa (~1/3 chance this year). And then you hope that your employer will agree to apply for a Green Card on your behalf right away (which they don't have to; they can keep you on the short leash of H-1B for another six years). The processing of the Green Card application can also take a few years. And if you change jobs while your Green Card application is in process, it's voided -- you have to have your new employer re-apply from scratch. And if, at any point during all of these, your current status (OPT or H-1B) expires, you have to pack up and GTFO of the country.

So no, not nearly the same in the US.

If i know correctly, the difference is you can't get your green card in those 3 years, where as here in Canada you can get your PR in those 3 years.

A non-zero amount of people never get the chance to obtain an immigrant or dual-intent visa during those years and subsequently get kicked out of the country.

In Canada you can pretty easily get a work permit to stay longer and PR without jumping through 70 hoops and waiting years.

The US immigration system is much shittier.

Yikes. That's a pretty heady classification of myself and my coworkers. The funniest part about your response, though, is that you said we'd "do anything for whatever else Amazon has". Looking past your assertion that we're in some way desperate, I'd say yeah, the whatever else Amazon has is pretty nice. I'm by no means a kool-aid drinker, but working with really smart coworkers, having accomplished mentors throughout the company, and working on genuinely interesting problems are nice perks to have. Amazon has its warts no doubt, but that's true of any company this size.

You can have your beefs with the company, that's fine. But I'd avoid the wholesale characterization of over 200k people -- it's not a good look.

I saw an IT Manager job description for Amazon's offices near Chicago. It had items like "Must be able to stand 10 hours a day." What? A managerial position where you're standing all day? Where you have direct reports and meetings and such? No offices or cubes for management? Oh, ok.

Also nice way of saying, "50 hour weeks are the norm."

It's because not everyone gets so emotionally carried away by what the media love to make of Amazon. The company employs 230,000 people. But for every one of them, it's just a day job after all. The scrutiny is justified given the scale and impact of the company, but what you're doing amounts to nothing more than misguided condescension from a bystander with little sense of what it's really like.

Playing such behavior down like "So many thousands work there so it's alright" is a very negligent attitude IMO. Companies that treat their potential employees like this need to be called out and shamed.

I'm not saying Amazon shouldn't scrutinize who they recruit. Please read the post with an open mind and think for yourself whether they have crossed lines of sensibility or not.

"nothing more than misguided condescension from a bystander" - No one is judging Amazon for its x thousand employees. Only its hiring practices. Which is creepy and stupid. I don't need to know everything about the company or its 230K employees to judge their hiring practice.

You misunderstand my point; by "scrutiny," I mean the media coverage on + testimony of things that amazon is doing wrong. I am all for bigCos being subject to watchful eyes. My qualm is that you somehow see this very specific problem we've gathered here to discuss as an excuse to hold moral high ground over the thousands of people who works for the company in a wake-up-sheeple style rant.

Init sure why you think this has anything to do with the people who are already working for the company. My comment was only about their hiring practice and potential candidates.

I do not hold moral high ground over the folks that made Amazon this attractive a tech place to work in the first place.

Still, just calling out their "current" hiring practice as evident from recent experience posts as creepy, stupid and something to be ashamed of.

I am amazed at how far you tolerated it. This is probably how I would have handled it:

> As preamble, the proctor made me download some software, one of which spun up a UI for chatting with the proctor and giving them access to my machine so they can take control of my entire computer, including mouse.

Nope. Goodbye.

Well it's good that someone tolerated it; now I have another couple names for my job black list. Thanks! :)

(Although I wouldn't work for Amazon anyway, but now I don't have to do any research if a company brings up "ProctorU").

You have to really watch those interview red flags. They're super important.

I had one right out of college where the office (on energy efficient lightning) looked like some cubicle version of hell. A person in the elevator on the way to the interview mouthed "run away" to me, and really creeped me out.

I got the job, but literally couldn't even handle a full day in that cave-like godawful place.

Holy cow.. I've got an Amazon SRE headhunter sitting in my inbox right now. I am sorely tempted to decline and send a link to this article as reason why.

I think that would be good feedback, if you're turning it down --hopefully it will incentivize them to change. If you're taking it, of course not.

Same here. It was for lab126, I was excited about it until I read this article..

Some other commenter up there says he worked at lab126 and it's sane there.

Save the link, you'll use it again and again.

I bit at an Amazon posting 5 years ago, but declined to do the Seattle interview when a local (and more attractive) position opened up.

Despite repeatedly telling Amazon's headhunters that I'm not interested, I hear from them about every 6 months.

Please do

Do it.

It gets worse -- when he wanted to leave the interview, the proctor couldn't disconnect those systems and kept making him wait until he finally gave up, cut off the call, and tried to purge it all himself.

That the electronic equivalent of preventing someone from leaving the building.

How else they would buy themselves the time to steal all his porn? ;-)

I would probably have tolerated it for a bit longer than that (On an old laptop or something). Up until the 'Clean your desk'.

Nope, Goodbye.

The clean your desk stuff is fine, but all of this should have been sent well ahead of the interview. Why they waste the time of two people to get ready for the interview is beyond me.

Yeah, I agree. Might have been willing to go through all that hassle. IF they had warned ahead of time the restrictions it might have been acceptable. I would pick another location, like the dining room or something, instead of having to 'clean' the desk. I tend to have an organized mess on my desk, and if I have to 'clean' it will drive me nuts for weeks. But, without heads up that is such a waste of time for everyone involved its crazy. And that's why it was the turning point for me.

"Hold on a second (while I spin up this virtual machine to install this software on...)."

That's when you should just start taking off your clothes just to make sure you're in even more compliance. (Might be writing on clothes)

I've got a server in my basement running VMware Workstation. That's how id approach it.

They won't let you continue if their software detects a virtualized system. Source: I used ProctorU for remote exams with Georgia Tech.

I feel in that case, it was acceptable. Students were encouraged to get a cheap "burner" windows machine and use it only for the proctored tests, and we were warned in advance of all the restrictions.

Seems a bit unreasonable to ask students (who likely already have tons of debt) to purchase an additional laptop just to take exams.

Yes, it does. On the other hand, though, you can get a cheap Windows laptop for less than some college text books.

Well, that's not saying very much, considering the whole thing with college text books. My last gaming rig literally cost less than my wife's textbooks last year.

You might be a bit out of date here. Textbooks are absurdly expensive these days relative to computers; Stewart's "Calculus" is $289[0], for example. That's about the right range for a new low end laptop or a used mid-range laptop.

[0] https://www.amazon.com/Calculus-James-Stewart/dp/1285740629/

Perhaps. But 1) you can rent that book for a fraction of the price (Amazon currently shows it for ~$60) and 2) just because students have to pay for expensive books doesn't mean they have extra money to spend on a second computer.

Some universities loan students laptops for free for like a week.

Oh yeah that's standard procedure for an interview, "I got an interview this afternoon, better build a new machine"

Yeah as a student they gave us about a month warning.

Seriously? Was this a professional program or something? Maybe I was a particularly poor undergrad, but I can't imagine being able to afford a separate computer, even a shitty one, with a month's morning. Nor could I justify the expense -- "oh that's just my test-taking computer."

It didn't have to run any real software except a browser and maybe a PDF reader. Also, yeah it was a remote grad program, so it was a bit of an experiment with how you do testing remotely at scale. I think they were trying to reach a fair compromise that didn't involve rewriting course content to include more open exams.

If you could use a virtualbox I would be totally ok with this, the fact that they exclude exactly that makes it 10 fold shittier.

If you consider the purpose is to monitor the entire computer to see if you have notes pulled up, you can see why they have this requirement.

I thought the entire premise of proctorU and proctorTrack (which gatech is now using instead of proctorU) was ridiculous though. A cheap hdmi splitter and a long cable would be all you needed to pipe the test out to a second monitor in another room where someone could scoop up the whole test for later. Or maybe even feed back answers to some tiny headphones hidden behind the ear. Impossible to detect.

Pretty sure that guaranteed-detection of virtualization is currently an unsolved problem, or we'd be hearing about breakthroughs on the "is the universe a simulation?" question.

In principle you are correct; in practice the commonly-available emulation environments don't do anything to hide the fact that they're emulations. If you ask the OS for the name of the graphics card and get "VirtualBox Graphics Adapter for Vista and Windows 7"...

I'm sure your VM doesn't have a webcam though. How do you prove that you are not cheating like this?

You can pass a USB device to the Hypervisor with nomachine then pass it to the VM with VMware.

You're gonna do all that work just so that you can take an hour long interview for an Amazon internship?

What work? It is just a VMWare knob.

just what are you to do if you don't have a webcam? go buy one?

Yup. You can get a camera that is good enough with a built in mic for like $20 these days.

My thoughts exactly. If I was reeeeally invested in getting that job, I would have found another laptop lying around, and powerwashed it after I was done.


Applications are open for YC Summer 2019

Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact