Caltech's CS interview prep course

<a href="https:&#x2F;&#x2F;archive.is&#x2F;njEHH" rel="nofollow">https:&#x2F;&#x2F;archive.is&#x2F;njEHH</a>

My career path in tech so far has been junior engineer -&gt; mid level -&gt; senior -&gt; tech lead.<p>As I&#x27;ve progressed through that path, I have had literally no opportunities of applying the stuff I&#x27;ve learnt on HackerRank, save for a couple of interviews.<p>At this point I&#x27;m convinced that the focus of those interviews was entirely wrong. Once you pass those tests and get hired, you will be dealing with tons of legacy code riddled with dumb queries, questionable code quality, and very often the wrong stack for the problem at hand.<p>I&#x27;d be OK with modern interviews if they reflected the job you will be doing to a reasonable degree. But knowing the state of the industry I feel that they fall somewhere between gatekeeping and a bait &amp; switch scheme.<p>If you haven&#x27;t spent enough time practising those problems, you will look like an idiot. If you have spent enough time practising them, your expectations will be high, and you will feel scammed when you read the steaming shit your new employer wants you to improve.

So wait you are smart enough to get into caltech, smart enough to graduate with a degree in CS. And yet the curriculum on its own does not do a good job enough on its own to prepare you for getting a job? Very hard to graduate with a decent GPA in CS at caltech and not be able to thrive as an entry level engineer. If you got B’s in caltech CS I’d make you an offer for 200k+ without blinking or even asking you anything.<p>I have a small friend circle, I would say they’re all smarter than I am on raw problem solving and intellect capability. They came out with better degrees from better institutions. All have PhDs and yet they’re all facing delays when landing a job. There is something definitely broken. I think the industry has created this problem and the industry should fix it.

No clue if this is even remotely true, but here is what I think: the most successful companies can basically pick and chose from a really large set of applicants. Given this luxury, the can, in essence, filter by any kind of metric they want and still get loads of suitable people.<p>I short: if you have a million applications you can sort by GPA and still have a hundred people to chose from. Sorting by GPA means you will overlook lots of raw talent, but who cares, there is still enough real talent in your funnel as well.<p>For me, someone hiring for a way smaller and poorer company, this is actually good. These processes overlook lots of great people, which we can snag up.<p>This observation is based on the behavior of big companies in Germany (think carmakers and similar), not sure how well it translates to the silicon valley situation, but seems similar.

If you don&#x27;t think modern interviewing practices are a problem, this should be proof enough. Not only do we have an entire programming book sub-genre dedicated to this nonsense, but now we&#x27;re also wasting class time on memorizing sorting algorithms. Fantastic.<p>Of course, we&#x27;re going to see the inevitable arguments that &quot;40% of the people I interview lie on resumes&quot; or &quot;I interviewed this guy once, and he couldn&#x27;t even write one line of code&quot;—let me pre-empt that by saying if you ever get to that stage, your screening process must absolutely suck and it&#x27;s still your fault. I&#x27;ve interviewed dozens of people and never had that experience, because I check OSS contributions, professional history, published material, and ask for code samples before I even engage with a candidate in a technical discussion. It&#x27;s not rocket science.<p>I&#x27;ve told at least 5 or 6 companies that wanted to white-board me to suck it. It&#x27;s honestly insulting that I&#x27;m forced to white-board when I have two books with my name on them (published by Apress), contributions to Golang (small commits, but they still got me in the AUTHORS file), and OSS projects with hundreds of stars on GitHub.

For context, this is a CS11 class which is a catch-all small form factor elective (3 units, average class is 9 units, units are supposed to be that many hours per week) that is used to offer a dozen or so small topics that are not extensive enough to warrant a full class. In any given term they will usually offer various CS11s for learning new programming languages and a few other random topics with some being run by grad or undergrad students.

Sometime around my Junior year of college my goals shifted from &quot;I&#x27;m here because I want to learn&quot; to &quot;I&#x27;m here because I want a job&quot;. I was surprised to find that doing well at learning (getting As) not only wasn&#x27;t good enough for an entry level job, it wasn&#x27;t even really correlated. The more time I spent on coursework the less time I had to grind through interview questions. I wonder if there are any other industries like this.

Are there other major economic sectors which have circled the wagons around making job applicants do remedial coursework as live performance art?

Interviewing is broken.<p>When I was at Amazon, it became impossible to hire candidates, and generally the engineers we brought on board were non-US citizens. Typically they&#x27;re Indians without U.S. residency status, often times U.S. college grads or often even transfers from Amazon&#x27;s Indian offices. Not only did we pay below other FAANGs, but with these hires, we&#x27;d pay them at the low end of our own scale.<p>What are they going to do? Quit? Ha! If they quit, they have to leave the country. So most of them just suck it up. Many of them had spouses and children. Quitting Amazon meant they&#x27;d have to uproot their entire family.<p>A second order effect the entire industry on how to work around Amazon (and other big tech&#x27;s) styling of interviewing. What amazes me are all the Youtube channels from Amazon engineers and other big tech companies with this &quot;influencer&quot; persona.<p>At the end of the day, many of these kids at Amazon are doing shitty migration projects, shoveling legacy code around all day, and dealing with on-call hell.<p>But we have this entire cult of working in big tech and going through all these gymnastics to get through pointless interviews.<p>Perhaps Google, Netflix, etc. are much better places to work, but it boggles my mind anyone would invest so much energy in getting a job at Amazon. It&#x27;s absolutely a shit place to work that doesn&#x27;t treat people like human beings.

LC interviews are well past the curve of utility. 10 years ago, FB and others would do a simple fizzbuzz + lc easy&#x2F;medium demonstrating that you had a good grasp of complexity, pointers, recursion, and iteration. I got a job or two by showing that I could DFS a binary search tree, count letter occurrences in a paragraph, or return the top k elements from two sorted lists.<p>Then people learned to practice these simple coding questions, and employers had to up the anty. Now everyone wants an LC expert who can do an LC hard in under thirty minutes with perfect code and no hinting. I wouldn&#x27;t be surprised if LC has to open a new category of problem difficulty based on current trend lines.<p>At some point you really are hiring for a skill which is only applicable in interviews. Solving an LC hard in under 30 minutes has little bearing on whether a candidate could solve a novel research problem or difficult refactoring task in their day to day work.

Other than the course outline, is there any substantive content available?<p>I saw only a link to the first problem set, and even that requires a Caltech login.

Tech interview discussions always seem to run the same course and hit the same points, so I&#x27;ll repeat one of mine:<p>What&#x27;s wrong with the idea of turning the standard Leetcode algorithms interview into a license that only needs to be completed once, or maybe once every 5-8 years, similar to what physicians must undergo? Much more efficient than having to retake it with <i>every single company</i> during an interview cycle.

Last time I interviewed, I thought about how it probably would have benefitted me to take one less class a semester in college and just drill leetcode-style questions in that time. Given how almost all &quot;prestigious&quot; companies use these problems heavily for interviewing, it&#x27;s likely to increase earnings more than any other course.<p>I think what caltech is doing here makes a lot of sense and demonstrates an awareness of the realities of interviewing at companies a lot of their students will want to work at. College is also a better time to build up this skill than later when other responsibilities tend to pile up.

Idea: Professional CS interviewers!<p>I&#x27;m a good programmer, but not a good interviewer. That makes me pretty typical.<p>I could spend a lot of time getting better at it, but I would only marginally improve, and it would take time out of the work I&#x27;m actually good at.<p>This feels like a typical problem for division of labor to solve.<p>Maybe it&#x27;s already done in done form?

I feel that the interview format heavily already favors new grads who just spent an entire year or more focused on algorithms courses and learning a lot of the theory you might get pop quizzed on.

This is absolutely embarrassing for the industry.

I understand part of the point of posting this is to demonstrate the problems with the educational system for CS majors and the hiring practices of the industry, but what is the actual point of sharing it here if the resources aren&#x27;t open for all of us?

Isn&#x27;t this the sort of thing university is not meant for? In Canada we separate out college and university, and this would mayyybe make sense as a college level course maybe, since it is more geared towards trades. But this sort of thing just reeks of corporate involvement in university and education of future scientists. It&#x27;s quite clear to me that this course is purely dictated by &quot;interviewing style of the day&quot; and is not really pushing forward the knowledge of humanity AT ALL. What is next? AWS console UI programming 101? That is not transferable knowledge is it...

I&#x27;m a bit confused. The course number on the page and in the URL is CS 11. But the Caltech course catalog lists CS 11 as &quot;Computer Language Lab&quot; [1] and the description is nothing like the CS 11 at the submitted URL.<p>[1] <a href="https:&#x2F;&#x2F;catalog.caltech.edu&#x2F;current&#x2F;courses&#x2F;department&#x2F;CS&#x2F;2021-22" rel="nofollow">https:&#x2F;&#x2F;catalog.caltech.edu&#x2F;current&#x2F;courses&#x2F;department&#x2F;CS&#x2F;20...</a>

Any way to access the assignments without a Caltech login?

I&#x27;m just dumbfounded that an university that has this:<p><a href="https:&#x2F;&#x2F;www.caltech.edu&#x2F;about&#x2F;legacy&#x2F;awards-and-honors&#x2F;nobel-laureates" rel="nofollow">https:&#x2F;&#x2F;www.caltech.edu&#x2F;about&#x2F;legacy&#x2F;awards-and-honors&#x2F;nobel...</a><p>...also has to offer a whole course to prep for job interviews :-|

This is a standard second semester Freshman year intro to programming curriculum. Most Universities have a class that looks exactly like this. I guess the only thing unique here is that when you&#x27;re ready to graduate, you can take a class you already took 3 years earlier.

all this jumping through hoop bullshit needs to stop. just take a genetic scan at birth, assign someone their lot in life, no more worries.

Wait I&#x27;m confused what happened that this link sparked so much discussion about how interviews are broken?

The link looks awesome, but is it just a listing? I couldn&#x27;t find any of the content

Could anyone kindly provide the materials associated with the course? Thank you!

I think the thing going on with interviews is that (most) of software engineering is actually pretty easy. Anyone can do the job outside of an interview session with the right experience.<p>Note how I stressed the word (most).<p>The interviews mostly serve to keep salaries high.

Thanks could not get main site to load. This should be at top.