@@spankyjeffro5320I had the same issue, but you can get yourself a nice pen. The pilot G-2 0.7 ones are pretty good. I had the same issue but had to get used to it for my math lectures. Didn't have money for a tablet :/
I feel it's a shame that people now think, I need to get better at programming, and they don't write some cool programs to become better but they will do leetcode problems for countless hours.
Sometimes it feels like the projects I do are too straightforward for me to actually learn as much as I want to. And the projects that look like they can teach me are often too high-level for me to even know where to start from. It's so confusing.
It is a slippery slope even for senior software engineers that simply are not used in the new shit, they really second guess themselves until they talk to someone they know (me) and shittalk them out of it, like... WWWWWWTTTTFFF you know what you can do etc, no you dont know everything, newsflash that's why we google things :p
@@datboi1861 I believe that's a pretty normal experience at first, but solving this by programming solutions will learn you a lot in the end, even with some projects that might appear simple at first.
I use leetcode for de-dusting my skills on a language. If I haven't used a given language for over 3-4 years, a dozen problems on all difficulties is more than enough to refresh for interviews.
There is a reason why he is very much into Leetcode, coz in India interview process is fucked up like anything, College Gard has to go through 4-6 rounds on interview in order to get internship at good product based company, hence he was compelled to do that much of programming, I don't blame him, on top that the job market is way more competitive in India.
Lmao don't bother about these folks, most of them would struggle to clear OAs and Rounds in Indian settings. Better to stick to what gets one a good job.
@@pratikwankhede6218 lmao, thing is dsa is not even useful in most jobs unless you're in database company or something lol. He's not wrong, the indian system is fucked up.
Indian interview changes once you aquired some industry experience. Then you will mostly be asked about the tech stacks and take home project is also very popular.
How the fuck can this be true with all the code monkeys Tata Consulting Services sends us in Europe? The worst code I have ever seen anywhere in my career is from big TCS projects . assertTrue(true) 🤯 Tests that only test auto injected mocks, etc. Do all the 1337 programmers leave for San Francisco?
Doing projects is a different way of thinking than solving a problem/question. I'd say doing projects teaches you the proper way of thinking when it comes to programming, while problems are just short brain exercises.
i have done many project and now i am back to leetcode lmao in the end grinding leetcode is required for coding interview prep sadly, most decent jobs got a white boarding session
To learn French you need to read, write, listen and speak to get proficient. To become a boxer you need to do sparring, drills, running and weight lifting. And to become an excellent programmer, depending on your domain, you might need to build projects, solve Leetcode problems and study mathematics. You always need a variety of stimulus to become proficient at something.
Yes, we need both but I think starting out focused mostly on Leetcode and codewars is a mistake. @@tapwater424 NOW that I have started building things I can do codewars better now.
0:22: 💡 Practicing interviews at companies you don't care about can help improve your skills and discover new interests. 3:02: 🔑 Having a basic understanding of fundamental data structures and algorithms, such as binary search, is crucial before attempting coding problems. 6:22: 💡 The video discusses resources and strategies for preparing for coding interviews. 9:17: ✏ Drawing and thinking is a helpful approach to problem-solving, and using a whiteboard can simulate an interview scenario. 12:16: 🔑 Staying organized and learning from others' solutions are key to success on LeetCode. 15:23: 📝 The video discusses the benefits and drawbacks of using Obsidian and Notion for tracking progress and organization. 18:25: 👍 The video provides valuable advice for coding interviews and emphasizes the importance of timed practice and stress-inducing environments. those are some good advices
You've no idea how many things I like to start at with the hardest difficulty, just because I didn't want to do anything less. Like Halo Combat Evolved.
@@nullx2368 trick is that you have to train to solve them for while. They are very similar to competitive programming and test your familiarity with this class of problems. Not something you can learn being a good software developer. So i suspect those trying to create sunken cost fallacy to give HR leverage.
One really nice thing about leetcode problems is that you can jump into one and finish it with very little warmup. On the flipside, if I'm working on a project, I need more time to get into a flow state in order to be productive. I can just crank out leetcodes because they are self-contained problems.
And this is a big problem. Because projects allow you to learn skills that is actually needed on the job: compiler optimization, caching, vectorization, threading, or even just setting up a build system. These would NEVER appear in leetcode and it’s a shame.
Solving leetcode or algoexpert problems is fun and challenging and will make you a better PROGRAMMER for sure, but does nothing for your DEVELOPER career imho. Being a good, professional developer is so much more than being able to sort a list or create a linked list or whatever. It's design patterns, using frameworks, working in a team, using GIT or some other source control system, etc. etc.
I think in companies, it is more about understanding frameworks, storage solutions, networking. the maximum what I had was if I understand the complexity O-Notation of a given problem.
The secret handshake of getting in the door is to solve DSA problems. The system is flawed, but wouldn't you rather just do it if it helps you get the job?
@@Stasisdrone4827but then when you do get the job and all you did was leetcode and never bothered to learn fundamentals like OS, networking etc, you'll gonna suck so bad in the industry.
They will teach you that Gabe but as the Ist commenter said dsa gets you through the door. Ofcourse when you get the experience your are expected to know things@@GabeTheYabe
"Apply to companies you don't care about." may not always be a good approach because not all companies have the same interview process. I have never in my life had a whiteboard interview so no matter how many interviews in smaller companies I did, I would never get anything near passing the big tech interview.
As a senior, I can guarantee you even you understand DSA fundamentals very well, you would easily fail at LeetCode questions, you just have to come back to LeetCode again and again. The problem is LeetCode skills don't stick.
I think I'm gonna start telling non-coders to tell me the hardest math they've ever done. Then I'm going to give them a time-limit and constantly pester them to tell me to explain their process. Then I'll judge them entirely on those 30 minutes, ignoring their work history and ability to explain their career beforehand. But first I'll make sure that my potential hires have autism to some degree, by ensuring it's a job with technical skills. Then instead of being sued for ADA violations, I'll become the standard, and force other companies to follow. Then I'll get youtubers like this guy to be so into it, he calls people who can't do it "dumb", regardless of how much code they've written and how clean it was. John Carmack's Quake code, for example, was OOP based, so clean a beginner could read it, and came out before leetcode... but that makes him dumb, obviously, because it didn't have this interview to prove himself, and likely would have failed. Q-bert was written by a 13 yr old. He too would've failed. Linux was written by some college kid.
The irony is no one is stopping the unemployed programmer from creating the next Quake, the next Linux blah blah... also John Carmack, Linus and the kid would've crushed leetcode, if that was their thing.
this is simply untrue. this just shows you are not a good problem solver at all. good competitive programmers never need to keep revising or memorizing shit , especially most lc problems are easy and not even close to implementation heavy
A flaw to the advice of applying at companies you dont care about to practice coding interviews is you assume we can even get an interview in the first place
As a junior/mid I'm on stackoverflow all the time. It usually points me in a good direction towards solving the problem I have at that moment... but I've never experienced situation where I was able to copy/paste some portion of the code that would just work. I suppose it's funny as a meme but I don't find it as a realistic situation.
@@o1-preview it actually has a completely different purpose. People who invest themselves enough to get good enough at leetcode and then go through their extremely lengthy process to work for them are more likely to stick with them for longer, and happy about working for these companies. The entire situation where a lot of people look at these companies as their career goal is created and maintained by this ecosystem where there are learning platforms, contests, courses, youtube channels and whatnot. Another thing they really try hard is pushing for not needing college or masters education for roles. The more people they can add to an industry that has way more role gaps than engineers, the lower the salaries and benefits get...
To be honest. I mostly experienced the opposite. In many cases the stack overflow solution isn't even a good direction to go in. Now I mostly look at source code because documentation is often outdated or incomplete and I test a bit around external functions to look if they actually work as expected or have some hidden bug. Especially apis are a bit crazy how many errors you can encounter that just aren't or false documented. Of course documentation is my first goto but if I can't figure it out with the documentation source code is the way to go. Edit: This might be a problem with the javascript ecosystem.
In our undergraduate studies in France, we have whiteboard sessions when a teacher asked us to solve stuff and roast us when we fail. Then grade us on the perf. You hear stuff like "why are you even here", "in one semester you are out for sure", "this is high school level math, and you still got it wrong". We were going in this hour of joy in group of 3 students and while you were solving your problem in one whiteboard in the corner of the room, you mate was being roosted by the professor. My favorite time was when my friend was done with the theory question and explain everything to the professor and the latter replied: "good, now you are going to demonstrate to me why what you just said is wrong". Don't miss it but if coding interviews are like that as well, the trick is to take it with a smile and don't pressure yourself too much, show confidence even if you are not sure how to solve and always try something
I've done all of the problems on leetcode. I just find it a fun thing to do when I'm watching a movie or waiting for something. Look, it's either sudoku or leetcode.
My biggest problem with leetcode is that I can never understand the problem! Whoever writes the descriptions for the problems doesn't write them nearly stupid enough for me to understand. It's a skill issue; if I knew what the answer was, I'd understand the question, but I'm still going to complain about it.
I had an interviewer paste a huge screenplay in the zoom chat of the happy number problem. I never have done LeetCode and have been a dev for 12 years. I have never got a LeetCode question in an interview until 2 months ago. I could not understand the problem at all. I said I dont understand the problem. With things like this I need time to myself to let the problem sink in. So he said ok lets do it as a take home and you finish it in your own time and email me back with the solution. So I did. I did some research and understood it and had a solution. He asked follow up questions and then I get rejected. I will from now on reject companies that do LeetCode questions OR if they do Codility/TestDome/HackerRank. Nope.
Lmao dont know if anyone has this issue like me but in some problems i got to attempt to solve it first to really understand the question and this makes me spend like over 30 minutes for them.
I will say it's not easy solving algos .. with experience and taking the time to learn the underlying concepts of the different algos you'll see. Solving algos becomes second nature because it intuitively makes sense to you what you're trying to do.
The important thing is: you need to be doing it regularly. In the past I also cracked several medium, hard problems, but then, as I never looked to them again, I just forgot about them. So if you are not doing it on a regular basis, your mind will erase it. Unless you are gifted by the universe with a freaking good memory (not my case at all).
If you are trying to remember problems, then it's temporary & not a skill. Do you need to learn to drive if you don't drive for an year or so ? Definitely not.
If you have to drive after a year and you are participating in a race(45 min interview) you definitely need to practice driving again. Atleast I need to do it, and never mind I agree, I am a looser, just not somebody making vedio (ranting) on youtube for no reason
@@liquidmetal718you’re not taking a test to drive, which is autopilot. You’re competing for a difficult driving test requiring full focus and being perfect in performance since the bar for passing is extremely high. You need to be at the top of your game because you need to do it with speed perfectly.
After 20 years, I can count how many times I've had to write a red-black tree in real life on 0 hands. If you have to implement this stuff yourself then you're doing something wrong.
I could maybe see it if you're working on some super-resource-constrained embedded project, and can hyper-optimize for your exact use case... Otherwise? Yeah, just use a library.
I agree very much, but I have no clue how I'd find anyone who I'm confident can actually write maintainable, reliable, testable code. It makes sense to give several Easy LC problems to verify they aren't faking being a programmer, rather than giving LC Hard problems to try to find a "10X" super-genius-super-special impossible-to-work-with engineer.
@@SolarShadoi work in embedded and honestly, there is very little usage of fancy algorithms, other than numerical domain-specific ones (FFT, signal processing algos, etc). I mainly write C in environments where dynamic memory allocation is prohibited, which makes 99% of cool data structures impossible or not worth it. I don’t think leetcode is very useful for embedded or systems programming, but maybe there are some good collections of problems for those domains im not aware of.
@@SolarShadoBut 99% don't and that's the point. The whole problem with leet code is that only pertains to people that already would know these things. Interviews that rely on leet code type of questions to determine a candidate's abilities are flawed from the get-go.
Ngl the interviewers in india for amazon google and meta asked me real hard questions for dsa almost like they wanted me to fail. I joined meta in london and the interview was much easier especially the dsa bit, also interviewers were more friendly and actually felt like they wanted me to join. I donno if this is universal or just something I felt.
if there are 1m people for your position, you better be 1 in 1m to get the position. once you get there its hard to lose your job but people get bored because they are 1 in 1m.
its cuz of the competition. Most of these interview rounds act as a filter, especially in places like India, where there are far more applicants for a role.
@@career800 oh yeah meta in 2021 mid was hiring like crazy I applied for a role in Dublin but it was an sre/dev role but mid way they told me a role in london was more suitable. I just applied randomly. Similarly had got offers from intercom and another unicorn during the time they were sponsoring visas for folks with around 4 years of xp
I'm so glad I live in a country where leetcode is not relevant and isn't asked at web development job interviews. I'm a web developer, I don't know what a binary search algorithm is and I don't need to know it xD.
It is similar to working in higher level languages making you better at programming in lower level languages. Knowing about leet techniques can make your normal code better.
Look whatever he said fits in indian scenario. I hope is just casual roast. To crack big MNC in india you need to be an expert/cm on codeforces specially in current market scenario as a fresh college grad
I have solved one leetcode problem so far. I want to solve about 10-20 but only because I want to see if I can do that in BASIC since I don't think anyone else does it. I kind of like many of the problems, in the same way I like solving crossword puzzles on the beach. Anyway, I think Rosetta code is more interesting though not quite equivalent.
Never in my life did I solved any of those coding problems. But I developed a huge IoT System for a german Super Market Brand, developed a PACS with a fancy 3D Voxel Engine for Medical Imaging. Never got a chance to work for FANG or Microsoft. I think I would fail at their tests badly. But am i a bad programmer? sure not. Since 4 years I'm a manager of a Business unit with 30+ developers....
I learnt programming cause it was fun, i used to build fun projects that came to my mind . i never thought programming would become this, big tech runined programming for younger generations
Being an Indian I can see where this rat race is coming from, here we are nurtured as bookworms and for all sorts of tests. So most of em fall into this trap instead of finding their own path, which in turn is the much easier option!
Y'all need to chill and reproduce less my niggah, you can't live a good life with so much competition. Africa's next => the competition there will be 10x India and Salary will be 10 times less than India.
If you have five interviews and two of them are coding, you might spend some time on other topics besides leetcode - design, project management, leadership, quality assurance, etc.
14:00 I find more and more I end up more confused leaving stack overflow than when I went in. A lot of people either marked it solved with "I solved it myself" with no solution, the classic "Duplicate of" and the duplicate doesn't solve the issue, or one question with 20 different solutions which are so different from each other that attempting all of them would waste my time.
CS graded uni in 1990 learnt binary search. Retiring next year, never once implemented my own b search. I did see just 6:38 once “a deadly embrace” with db table locks.
Being good at leetcode tends to lead to shock and awe when the leetcode pro comes in on day one and is trying to work through legacy code with zero documentation. Best interview I ever had was a company that gave me a sample app where I had to 1: get 3 failing tests to pass 2: add a feature 3: suggest ways to make the app more efficient This is a much better approach that is more realistic for what you are going to do in your day to day. And the final suggestion piece can weed out those that cannot communicate well. Done.
Summary/ advice: 1. 7:46 do a good data struct/algorithms course -understand fundamentals -do 50 leet problems (easy-medium), targeting Arrays[1] -get some DP experience go to interviews with that sweet DP experience (DP = dynamic programming) 2. solve problems on with pencil and paper first talking out loud how it would work (imitating a white board) go 3. 19:09 try getting 3 commits in a opensource that you are really interested in
Thanks for the advice. Going through tech test hell now. Also totally agree with Geek for Geeks it is so horrible with all the popups. Even desperate for an answer I never click on Geek for Geeks.
Ive been a php programmer for many years, and dont know anything about sorting, an O and all that. Its not really needed in PHP, but im preparing to expand my knowledge, and other languages currently.
You can verify that he did the 1583 Leetcode questions by how trash this editing is, I lost the point in the first 5 transitions and felt dizzy and sick
Agreed about Leetcode performance numbers. It would be ideal if they could sum up the cost of ops from an empirical algo point of view - not Big-O, but actual sum. As the same test cases are run for all, it would be a good relative benchmark.
I've implemented it once. I pulled in data from a crypto website and each coin/token started it's data on a different day. So I just picked the first day the company existed to check and then used binary search to find that first day where the data started flowing, for each token. So far the only time I've implemented it.
Leetcode trains a certain set of skills, that are usually not used by most professionals in their specific careers, that’s why a lot of people talk about it not sticking.
FYI: AVL Trees are balanced. Yes, AVL trees are balanced. An AVL (Adelson-Velsky and Landis) tree is a self-balancing binary search tree where the difference between heights of left and right subtrees cannot be more than one for all nodes. This property of AVL trees ensures that the tree remains balanced, leading to O(log n) search times, where n is the number of nodes in the tree. This makes AVL trees very efficient for lookup operations.. Source: Conversation with Bing, 18/11/2023 (1) en.wikipedia.org. en.wikipedia.org/wiki/AVL_tree. The example shown is unbalanced.
it is though: node 12's left subtree has height three, right has height two; node 18's left subtree has height one, right has height zero; node 8's left subtree has height two, the other has height one; and node 5 is the same as node 18 It doesn't have to be weight-balanced (i.e. roughly the same number of nodes in the left and right subtrees), only height-balanced
"The best way to to crack an interview is to give as many interviews as you can" - Life Lesson. 5 things you need to know that can save your life in coding interviews: 1. Hashing and hash maps 2. Binary Search 3. Tree traversal 4. Two pointer problems 5. Basic DP (top down and bottom up)
I found leetcode recently and it's a fun place that will throw out some puzzles. But it hasn't helped me build anything. I can make this one system REALLY efficient and consider how it could be optimized but... That's what you pull out after you've figured out how to build the whole system.
It makes you able to come up with your own, good solutions when the pre-existing ones doesn't deliver. Also makes you able to understand existing library code.
While i like stuff like leetcode and codinggame and etc, i feel like most times it ends up pushing people into "small domain" mindsets. Which is fine for "find the shortest path from x to y". But at some point I'll need you to "extend large domain x with new feature z. x is already a s**tshow of a system, how would you go about implementing z without making x even more of a s**show?". Now i might need an answer that totally runs counter to "best at the small domain" but is "best at the larger domain". Can the person do that shift if he has focused too tightly on the "small domain" issues? Can he understand that many times, there are tradeoffs? Has he become so familiar with the optimal solution that now he can't see he's trying to fit a square peg into a round hole?
For simple questions, I suppose it’s better to ask ai than searching anywhere else like stackoverflow. In 99 of 100 cases it perfectly gives you a simple example and brief explanation that’s enough to proceed further on your own. Another option is to look at docs, but sometimes it just takes a bit more time to find what you need than ask ai.
You also don't have to deal with some egomaniac being like "Simply... blah blah." and treating you like a dumbfuck for not finding a necro thread from 5 years ago on the same topic.
I just had an interview at a large investment banking company for an entry-level position. Throughout the interview, the interviewer kept saying, "If you don't pass this, you won't make it to the next round." Twice, he muted his microphone and turned off his video. While I was drawing on CoderPad's drawing pad, I was instructed to start coding. When I asked clarifying questions, I wasn't given any guidance, only "yes" or "no" responses. I usually visualize the problem first by drawing it, then write pseudocode, but I wasn't given the chance. I really wanted the job and was looking forward to the interview. It was disappointing, but now I just have to improve.
Stop gaslighting people. Binary search isn't easy at all. Quote from StackOverflow: Apart from Knuth's quote that "Although the basic idea of binary search is comparatively straightforward, the details can be surprisingly tricky", there is the staggering historical fact (see TAOCP, Volume 3, section 6.2.1) that binary search was first published in 1946 but the first published binary search without bugs was in 1962.
Agree, I think best thing one can do with binsearch on interview is to state that they can be some one off errors there and use example to make sure it's fine
"Before jumping into leetcode to get good at interview questions, make sure you actually have taken a class or two on coding or finish your degree". I'd hypothesize that actually sitting there in the interview thinking about the questions does slightly more than just solving leetcode problems. Not that you wouldn't encounter similar basic tests of your understanding, but just completing coding problems is different from what many interviewers want. I don't think they want just a guy who can solve their problem, they want a guy who demonstrates their process, the how and why and is good at explaining what they're doing and why. The why. That's where you really get tripped up about your knowledge, you might be technically breezing through problems, but it might just be that you don't know what you're doing and why. The employer usually wants you to know what you're doing when you do it even if you can't solve all the problems right away. And like I believe the big O questions are popular just because it demonstrates your understanding of what the impacts of your solutions are. I love the pen and paper point. For some reason (maybe obvious reasons) university notes and recommendations have moved from paper to virtual and I think it shows with my classmates. People don't necessarily write the notes anymore because the pdf is there, or when solving they don't start from the basics like understanding the situation and making a note on what they need to solve and what they need for solving it. They don't construct a plan (I'll admit I used to be really bad at that too and still am often too lazy, but I've reached a point where I know the subjects well enough to follow the intuition without getting tripped) or revisit theory/examples behind the topic if they're not sure how the phenomenon works. And I think it's a similar problem with the technical ability in coding interviews vs getting asked the tough questions like "have you thought about what you're doing?" I've recently started to check out a bit of design methodologies in manufacturing and trying to read the instructions through and framing the problem before attacking it. Just to not do mindless unnecessary work and get to the goal in a more coherent way. The biggest threat to that is if you can just about keep all the related concepts in your head about the problem so you won't make notes about a plan and then you forget something or get confused. And personally I'm really bad at evaluating my results before calling it done. I can often afterwards reason about them successfully to at least gauge how realistic it is, but throw it off the table as soon as I reach the answer. And that should be a priority skill. Definitely useful advice on this video if you apply them with thought, not all are relevant to just you, but plenty of good things just for having a structured process of learning. Which is the most difficult thing to have if optional, being organised.
The only time I've copy pasted from SO directly and it just worked was for this problem: the percentage of overlap between rectangle A and B, given that rectangle A is always bigger, and all we have is their bounding box coordinates. Someone happened to have the same exact question and there was a super elegant function that accepted two objects as arguments with the same name for keys. It felt like cheating and it has never happened again lol
I wish I could get that many interviews like he mentions in the beginning, I don't even get that many so that adds extra stress to the few that I do get, which have been very very few so far.
Learn data structures and programming fundamentals , do 50 leetcodes question easy and medium , get array styles and dp expirences and go for interviews , failing three interview is no different than doing thousand leetcode questions . use pen and paper , talk it loud like interview process using whiteboard . important talk through your thinking and do let know how you think . step 2 : after solving problem in 1 hour see other people solution , how its work what unique about that etc . aVOID BUSYNESS - DONT NEED TO record everything , need no to use notion , obsidian and extra apps for record everything and organize . stack overflow - skill issue , know when to use apis . can use timer to get interviewer view , and zone or can apply for leetcode contest for get in zone -- need that environment with people watching and timer . goal : leetcode in mediumi in 15 minutes , and how good u are at solving new problem . 3 /5 in contest . try to get good three open commit on open source . choose a technology and try it .
Only thing I ask for in interviews is to use good variable names. I've seen people use "start" and "start_time" in the same block - often even confusing themselves.
I'd add to use a language that has a lot of shorthand built in so you can code faster. Like Python and their list comprehension. Nice to be able to transpose a matrix in one line of code.
prime as a typical MANGA dumba is like: "NOOOOO, YA CANT CODE WITHOUT BEING ABLE TO CRACK HARD LEETCODE TASKS", and all the adequate 99% middles/seniors are like: "My code goes brrrrrr"
He has explained this multiple times. There is a secret handshake in tech that basic algorithms and data structures is what is required to get the job since it is a skillset everyone needs. If you dont want to do the secret handshake, good luck getting a dev job somewhere but it wont be at the top tech companies.
@@TJackson736 "skillset everyone needs" and "top tech companies" are hardly related to each other especially when we talk about algos and data structures. Only interns and database devs need them lol
imo if you want to actually be good you should learn the basics at the very least. you dont need to learn convex optimization or anything but leetcode pretty much covers those basics unless you want to stick to web dev code monkeying your whole career you will need these skills at some point tbf though i just described ~50-60% of FAANG devs. newbies ooh and aah at big tech but you really dont need to be good to work there
@@gianni50725 leetcode is so complex actually. You won't need to use algos irl in 99.99% of cases And web dev is the hardest field for now. Nowhere else you'll face such amount of complexity and different tools
More important, unless you're interviewing with a company where everyone speaks Hindenglish, learn proper pronunciation of the language that people speak at the workplace. Communication is important, and they might not tell you why you didn't get the job, but this could be the reason. Just knowing the words isn't enough, you have to make the correct mouth sounds.
"Go and interview at places you would never work at" Sure, I mean, it takes about 1,000 applications to get one interview, but we can waste two or three interviews just to get the practice in.
The entire system is designed to avoid having to hire Computer Science students. Which is actually fine. If you have a degree OR can prove proficiency through this type of interview, that helps out people who were denied a good formal education thanks to lousy K-12 schools. Everyone should have a path to catch up regardless of what zip code they were born in. There are plenty of jobs that are less demanding that are more about data where you don't need to know binary sorting off the top of your head. But you're going to struggle with business logic if you don't have a good analytical mind. A math degree is also very valuable in the tech industry.
True. No amount of leetcode can prep you. 4 years of experience, and I'm on my 5th company. Amount of interviews, idfk anymore. And yes, I still get the jitters.
Because anyone can just learn the majority of the skills on the job, but you can't learn those "on the job" specific skills if you aren't good at interviews, salary negotiations, or resume/portfolio prep.
"applying for the companies that you did not care about": worked till 2022. Now when everybody applies everywhere, you have no option anymore to interview "for free".
The biggest problem with pen and paper or whiteboards is that I need to actually remember the legacy skill of handwriting.
My handwriting looks like that of a child's. A contemporary "Eww, pencil? I use tablets for everything." child. XD
@@spankyjeffro5320I had the same issue, but you can get yourself a nice pen.
The pilot G-2 0.7 ones are pretty good.
I had the same issue but had to get used to it for my math lectures.
Didn't have money for a tablet :/
Always have a permanent marker, that way they know you’ve been there.
I struggle to write my signature
Handwriting was deprecated in life v.2008
I feel it's a shame that people now think, I need to get better at programming, and they don't write some cool programs to become better but they will do leetcode problems for countless hours.
It's a shame that people don't use quotation marks anymore.
Sometimes it feels like the projects I do are too straightforward for me to actually learn as much as I want to.
And the projects that look like they can teach me are often too high-level for me to even know where to start from.
It's so confusing.
It is a slippery slope even for senior software engineers that simply are not used in the new shit, they really second guess themselves until they talk to someone they know (me) and shittalk them out of it, like... WWWWWWTTTTFFF you know what you can do etc, no you dont know everything, newsflash that's why we google things :p
Yes, that's why I never applied at FAANG companies. Right from the get-go, I hated the Leetcode-esque interviews.
@@datboi1861 I believe that's a pretty normal experience at first, but solving this by programming solutions will learn you a lot in the end, even with some projects that might appear simple at first.
I use leetcode for de-dusting my skills on a language. If I haven't used a given language for over 3-4 years, a dozen problems on all difficulties is more than enough to refresh for interviews.
This is actually a great idea
what do you think of codewars?
There is a reason why he is very much into Leetcode, coz in India interview process is fucked up like anything, College Gard has to go through 4-6 rounds on interview in order to get internship at good product based company, hence he was compelled to do that much of programming, I don't blame him, on top that the job market is way more competitive in India.
Lmao don't bother about these folks, most of them would struggle to clear OAs and Rounds in Indian settings. Better to stick to what gets one a good job.
@@pratikwankhede6218 lmao, thing is dsa is not even useful in most jobs unless you're in database company or something lol. He's not wrong, the indian system is fucked up.
Indian interview changes once you aquired some industry experience. Then you will mostly be asked about the tech stacks and take home project is also very popular.
I would say the vice versa is more true
How the fuck can this be true with all the code monkeys Tata Consulting Services sends us in Europe? The worst code I have ever seen anywhere in my career is from big TCS projects . assertTrue(true) 🤯 Tests that only test auto injected mocks, etc.
Do all the 1337 programmers leave for San Francisco?
I stopped leetcode and started focusing 100% on building projects. Leetcode is good but I think starting out building things is completely diff
Doing projects is a different way of thinking than solving a problem/question.
I'd say doing projects teaches you the proper way of thinking when it comes to programming, while problems are just short brain exercises.
i have done many project and now i am back to leetcode lmao
in the end grinding leetcode is required for coding interview prep sadly, most decent jobs got a white boarding session
You need both but I think coders focus on nerd stuff to me. @@SealedKiller
To learn French you need to read, write, listen and speak to get proficient. To become a boxer you need to do sparring, drills, running and weight lifting. And to become an excellent programmer, depending on your domain, you might need to build projects, solve Leetcode problems and study mathematics. You always need a variety of stimulus to become proficient at something.
Yes, we need both but I think starting out focused mostly on Leetcode and codewars is a mistake. @@tapwater424 NOW that I have started building things I can do codewars better now.
Wow, start with easy problems and then increase difficulty. This is really next level advice. Life changing
0:22: 💡 Practicing interviews at companies you don't care about can help improve your skills and discover new interests.
3:02: 🔑 Having a basic understanding of fundamental data structures and algorithms, such as binary search, is crucial before attempting coding problems.
6:22: 💡 The video discusses resources and strategies for preparing for coding interviews.
9:17: ✏ Drawing and thinking is a helpful approach to problem-solving, and using a whiteboard can simulate an interview scenario.
12:16: 🔑 Staying organized and learning from others' solutions are key to success on LeetCode.
15:23: 📝 The video discusses the benefits and drawbacks of using Obsidian and Notion for tracking progress and organization.
18:25: 👍 The video provides valuable advice for coding interviews and emphasizes the importance of timed practice and stress-inducing environments.
those are some good advices
@@hanhthien2948c'mon dude, really? lol
@@hanhthien2948 which AI did you use to extract these video notes? if not, then it's impressive.
You've no idea how many things I like to start at with the hardest difficulty, just because I didn't want to do anything less. Like Halo Combat Evolved.
@@sunxsky5569 even if im 12 days late, he used HarpAI most likely
I would like to see prime doing a leetcode contest. it will be interesting to see how he approaches problems.
Bla bla bla
💀
He would likely fail at most of them. I would and I'm a senior... Most of these problems don't even help in real problems lol
@@nullx2368 trick is that you have to train to solve them for while. They are very similar to competitive programming and test your familiarity with this class of problems. Not something you can learn being a good software developer. So i suspect those trying to create sunken cost fallacy to give HR leverage.
leetcode is a skill in itself, i doubt he would be good at it
One really nice thing about leetcode problems is that you can jump into one and finish it with very little warmup. On the flipside, if I'm working on a project, I need more time to get into a flow state in order to be productive. I can just crank out leetcodes because they are self-contained problems.
And this is a big problem. Because projects allow you to learn skills that is actually needed on the job: compiler optimization, caching, vectorization, threading, or even just setting up a build system. These would NEVER appear in leetcode and it’s a shame.
Solving leetcode or algoexpert problems is fun and challenging and will make you a better PROGRAMMER for sure, but does nothing for your DEVELOPER career imho. Being a good, professional developer is so much more than being able to sort a list or create a linked list or whatever. It's design patterns, using frameworks, working in a team, using GIT or some other source control system, etc. etc.
I think in companies, it is more about understanding frameworks, storage solutions, networking. the maximum what I had was if I understand the complexity O-Notation of a given problem.
The secret handshake of getting in the door is to solve DSA problems. The system is flawed, but wouldn't you rather just do it if it helps you get the job?
you deal with the given cards @@Stasisdrone4827
@@Stasisdrone4827but then when you do get the job and all you did was leetcode and never bothered to learn fundamentals like OS, networking etc, you'll gonna suck so bad in the industry.
They will teach you that Gabe but as the Ist commenter said dsa gets you through the door. Ofcourse when you get the experience your are expected to know things@@GabeTheYabe
there are some couple ones where is important, embeddings is when you start to take care of what you write, btw I'm codeforces bucket
"Apply to companies you don't care about." may not always be a good approach because not all companies have the same interview process. I have never in my life had a whiteboard interview so no matter how many interviews in smaller companies I did, I would never get anything near passing the big tech interview.
im with the same employer for years, however i at least go to an interview 1-2 a year. its also good to get in touch with reality sometimes :)
As a senior, I can guarantee you even you understand DSA fundamentals very well, you would easily fail at LeetCode questions, you just have to come back to LeetCode again and again. The problem is LeetCode skills don't stick.
I think I'm gonna start telling non-coders to tell me the hardest math they've ever done. Then I'm going to give them a time-limit and constantly pester them to tell me to explain their process.
Then I'll judge them entirely on those 30 minutes, ignoring their work history and ability to explain their career beforehand.
But first I'll make sure that my potential hires have autism to some degree, by ensuring it's a job with technical skills.
Then instead of being sued for ADA violations, I'll become the standard, and force other companies to follow.
Then I'll get youtubers like this guy to be so into it, he calls people who can't do it "dumb", regardless of how much code they've written and how clean it was.
John Carmack's Quake code, for example, was OOP based, so clean a beginner could read it, and came out before leetcode... but that makes him dumb, obviously, because it didn't have this interview to prove himself, and likely would have failed.
Q-bert was written by a 13 yr old. He too would've failed.
Linux was written by some college kid.
The irony is no one is stopping the unemployed programmer from creating the next Quake, the next Linux blah blah... also John Carmack, Linus and the kid would've crushed leetcode, if that was their thing.
this is simply untrue. this just shows you are not a good problem solver at all. good competitive programmers never need to keep revising or memorizing shit , especially most lc problems are easy and not even close to implementation heavy
its because it short, like it doesnt stick at all, and it also contains alot of math stuff, like its not just coding
@@_Cfocus where is the math stuff? I've done over 400 problems
A flaw to the advice of applying at companies you dont care about to practice coding interviews is you assume we can even get an interview in the first place
As a junior/mid I'm on stackoverflow all the time. It usually points me in a good direction towards solving the problem I have at that moment... but I've never experienced situation where I was able to copy/paste some portion of the code that would just work. I suppose it's funny as a meme but I don't find it as a realistic situation.
trust me, no senior has to do that in real life either, the interview process is outdated asf
@@o1-preview it actually has a completely different purpose. People who invest themselves enough to get good enough at leetcode and then go through their extremely lengthy process to work for them are more likely to stick with them for longer, and happy about working for these companies. The entire situation where a lot of people look at these companies as their career goal is created and maintained by this ecosystem where there are learning platforms, contests, courses, youtube channels and whatnot. Another thing they really try hard is pushing for not needing college or masters education for roles. The more people they can add to an industry that has way more role gaps than engineers, the lower the salaries and benefits get...
To be honest. I mostly experienced the opposite. In many cases the stack overflow solution isn't even a good direction to go in.
Now I mostly look at source code because documentation is often outdated or incomplete and I test a bit around external functions to look if they actually work as expected or have some hidden bug.
Especially apis are a bit crazy how many errors you can encounter that just aren't or false documented.
Of course documentation is my first goto but if I can't figure it out with the documentation source code is the way to go.
Edit: This might be a problem with the javascript ecosystem.
In our undergraduate studies in France, we have whiteboard sessions when a teacher asked us to solve stuff and roast us when we fail. Then grade us on the perf.
You hear stuff like "why are you even here", "in one semester you are out for sure", "this is high school level math, and you still got it wrong". We were going in this hour of joy in group of 3 students and while you were solving your problem in one whiteboard in the corner of the room, you mate was being roosted by the professor. My favorite time was when my friend was done with the theory question and explain everything to the professor and the latter replied: "good, now you are going to demonstrate to me why what you just said is wrong". Don't miss it but if coding interviews are like that as well, the trick is to take it with a smile and don't pressure yourself too much, show confidence even if you are not sure how to solve and always try something
sounds a bit toxic, interviewers are super supportive
@@roshatron you must not have interviewed at Facebook then.
Lol, I had a professor just like that at my uni in Ukraine. Maybe a little less toxic though 😂
@@roshatronif this is ur mindset you'll never get a job
@@bobsemple9341 I already have a job and none of my interviewers roasted me when I couldn't solve a question.
Stack overflow is for finding out what the universe of solutions might look like.
I've done all of the problems on leetcode. I just find it a fun thing to do when I'm watching a movie or waiting for something. Look, it's either sudoku or leetcode.
My biggest problem with leetcode is that I can never understand the problem! Whoever writes the descriptions for the problems doesn't write them nearly stupid enough for me to understand. It's a skill issue; if I knew what the answer was, I'd understand the question, but I'm still going to complain about it.
Same.
I had an interviewer paste a huge screenplay in the zoom chat of the happy number problem. I never have done LeetCode and have been a dev for 12 years. I have never got a LeetCode question in an interview until 2 months ago. I could not understand the problem at all. I said I dont understand the problem. With things like this I need time to myself to let the problem sink in. So he said ok lets do it as a take home and you finish it in your own time and email me back with the solution. So I did. I did some research and understood it and had a solution. He asked follow up questions and then I get rejected. I will from now on reject companies that do LeetCode questions OR if they do Codility/TestDome/HackerRank. Nope.
Lmao dont know if anyone has this issue like me but in some problems i got to attempt to solve it first to really understand the question and this makes me spend like over 30 minutes for them.
Yeah actually I thought the same thing. The way they word the problem it just makes no sense whatsoever.
Maybe your English isn't strong enough?
I will say it's not easy solving algos .. with experience and taking the time to learn the underlying concepts of the different algos you'll see. Solving algos becomes second nature because it intuitively makes sense to you what you're trying to do.
The important thing is: you need to be doing it regularly. In the past I also cracked several medium, hard problems, but then, as I never looked to them again, I just forgot about them. So if you are not doing it on a regular basis, your mind will erase it. Unless you are gifted by the universe with a freaking good memory (not my case at all).
If you are trying to remember problems, then it's temporary & not a skill. Do you need to learn to drive if you don't drive for an year or so ? Definitely not.
If you have to drive after a year and you are participating in a race(45 min interview) you definitely need to practice driving again.
Atleast I need to do it, and never mind I agree, I am a looser, just not somebody making vedio (ranting) on youtube for no reason
@@liquidmetal718you’re not taking a test to drive, which is autopilot. You’re competing for a difficult driving test requiring full focus and being perfect in performance since the bar for passing is extremely high. You need to be at the top of your game because you need to do it with speed perfectly.
After 20 years, I can count how many times I've had to write a red-black tree in real life on 0 hands. If you have to implement this stuff yourself then you're doing something wrong.
I could maybe see it if you're working on some super-resource-constrained embedded project, and can hyper-optimize for your exact use case... Otherwise? Yeah, just use a library.
I agree very much, but I have no clue how I'd find anyone who I'm confident can actually write maintainable, reliable, testable code. It makes sense to give several Easy LC problems to verify they aren't faking being a programmer, rather than giving LC Hard problems to try to find a "10X" super-genius-super-special impossible-to-work-with engineer.
@@SolarShadoi work in embedded and honestly, there is very little usage of fancy algorithms, other than numerical domain-specific ones (FFT, signal processing algos, etc). I mainly write C in environments where dynamic memory allocation is prohibited, which makes 99% of cool data structures impossible or not worth it. I don’t think leetcode is very useful for embedded or systems programming, but maybe there are some good collections of problems for those domains im not aware of.
@@SolarShadoBut 99% don't and that's the point. The whole problem with leet code is that only pertains to people that already would know these things. Interviews that rely on leet code type of questions to determine a candidate's abilities are flawed from the get-go.
Same. Nearly 25 years here. Day to day, I always wind up having to handle strings up in the guts. Ideal? Hell no, reality? Yup.
the IRONY, my guy talking about DSA and then reprints everything from the start @4:30
Learn as I go is my approach here, reading a book about algorithms before even starting LeetCode is a good way to become demotivated
Ngl the interviewers in india for amazon google and meta asked me real hard questions for dsa almost like they wanted me to fail. I joined meta in london and the interview was much easier especially the dsa bit, also interviewers were more friendly and actually felt like they wanted me to join. I donno if this is universal or just something I felt.
if there are 1m people for your position, you better be 1 in 1m to get the position. once you get there its hard to lose your job but people get bored because they are 1 in 1m.
How did you even make it to Amazon london from india?
its cuz of the competition. Most of these interview rounds act as a filter, especially in places like India, where there are far more applicants for a role.
@@career800 oh yeah meta in 2021 mid was hiring like crazy I applied for a role in Dublin but it was an sre/dev role but mid way they told me a role in london was more suitable. I just applied randomly. Similarly had got offers from intercom and another unicorn during the time they were sponsoring visas for folks with around 4 years of xp
@@o1-previewNope, for companies you are just a cog, especially in India, which is why the salaries are low as there is always someone to fill in
I use stackoverflow not just as a quick reference but because it helps check if there's a new way of doing something.
You won't find "new ways" on stack
It's very rare to see a software engineer TH-camr like you... Keep it up🙌👏
How do you know other software engineer TH-camrs have a hard time keeping it up?
Cuz they NEVER do the hard problems
They cant hack it as real devs and become youtubers.
Prime explaining his notes brought back my interest in discrete maths lmao thanks
i solved 1476 problems on leetcode (Rank 2,270). But there's still type of problem patterns which i find hard to tackle, such as problem 432.
I am senior developer (7yrs) , I find leetcode hard.
I like how you have to implement binary search tree algorithm on an interview for a job for which you’ll do crud apps. Awesome idea :)
I'm so glad I live in a country where leetcode is not relevant and isn't asked at web development job interviews. I'm a web developer, I don't know what a binary search algorithm is and I don't need to know it xD.
9:27 "get the whiteboard out", did he just suggest..
bringing your own scrum mainer to an interview? CHROOT
It is similar to working in higher level languages making you better at programming in lower level languages. Knowing about leet techniques can make your normal code better.
Look whatever he said fits in indian scenario. I hope is just casual roast. To crack big MNC in india you need to be an expert/cm on codeforces specially in current market scenario as a fresh college grad
I have solved one leetcode problem so far. I want to solve about 10-20 but only because I want to see if I can do that in BASIC since I don't think anyone else does it.
I kind of like many of the problems, in the same way I like solving crossword puzzles on the beach. Anyway, I think Rosetta code is more interesting though not quite equivalent.
Never in my life did I solved any of those coding problems. But I developed a huge IoT System for a german Super Market Brand, developed a PACS with a fancy 3D Voxel Engine for Medical Imaging. Never got a chance to work for FANG or Microsoft. I think I would fail at their tests badly. But am i a bad programmer? sure not. Since 4 years I'm a manager of a Business unit with 30+ developers....
> "I kinda know where I'm good at, and I kinda know where I'm bad at"
> Camera fucking dies
OMG THE TIMING WAS JUST THE BEST
I learnt programming cause it was fun, i used to build fun projects that came to my mind . i never thought programming would become this, big tech runined programming for younger generations
Being an Indian I can see where this rat race is coming from, here we are nurtured as bookworms and for all sorts of tests. So most of em fall into this trap instead of finding their own path, which in turn is the much easier option!
Y'all need to chill and reproduce less my niggah, you can't live a good life with so much competition. Africa's next => the competition there will be 10x India and Salary will be 10 times less than India.
BASED
Wait, isn't knowing FizzBuzz enough?!
the bro culture is still in tech, just look at what most tech youtuber is talking about when they look in an interview.
today I learned that I am a loser.
Aren't we all
Today means everyday for me everytime I read someone else's code
You are not alone 😢
Me too 😭
Wining is another term for losing 1000 times trying.
If you have five interviews and two of them are coding, you might spend some time on other topics besides leetcode - design, project management, leadership, quality assurance, etc.
Bro where os your 9hr long algo course link i am not able to find it.
Same. Not in links
Search for "The Last Algorithms Course You'll Need" or just "primeagen dsa course "
@@carlerikkopseng7172Search it on TH-cam, you'll get it
It's on frontend masters but not free as of now. Course Name: The Last Algorithms Course You'll Want (part 1 & 2)
14:00 I find more and more I end up more confused leaving stack overflow than when I went in. A lot of people either marked it solved with "I solved it myself" with no solution, the classic "Duplicate of" and the duplicate doesn't solve the issue, or one question with 20 different solutions which are so different from each other that attempting all of them would waste my time.
CS graded uni in 1990 learnt binary search. Retiring next year, never once implemented my own b search.
I did see just 6:38 once “a deadly embrace” with db table locks.
agreed, and thank you
Being good at leetcode tends to lead to shock and awe when the leetcode pro comes in on day one and is trying to work through legacy code with zero documentation. Best interview I ever had was a company that gave me a sample app where I had to 1: get 3 failing tests to pass 2: add a feature 3: suggest ways to make the app more efficient
This is a much better approach that is more realistic for what you are going to do in your day to day. And the final suggestion piece can weed out those that cannot communicate well.
Done.
Summary/ advice:
1. 7:46 do a good data struct/algorithms course
-understand fundamentals
-do 50 leet problems (easy-medium), targeting Arrays[1]
-get some DP experience
go to interviews with that sweet DP experience
(DP = dynamic programming)
2. solve problems on with pencil and paper first talking out loud how it would work (imitating a white board)
go
3. 19:09 try getting 3 commits in a opensource that you are really interested in
Thanks for the advice. Going through tech test hell now. Also totally agree with Geek for Geeks it is so horrible with all the popups. Even desperate for an answer I never click on Geek for Geeks.
Ive been a php programmer for many years, and dont know anything about sorting, an O and all that. Its not really needed in PHP, but im preparing to expand my knowledge, and other languages currently.
Wow that's good.
I've heard php is a low tier language. No offence I'm just curious
Great idea about applying at places you don’t want to work at for interview practice.
You can verify that he did the 1583 Leetcode questions by how trash this editing is, I lost the point in the first 5 transitions and felt dizzy and sick
Agreed about Leetcode performance numbers. It would be ideal if they could sum up the cost of ops from an empirical algo point of view - not Big-O, but actual sum. As the same test cases are run for all, it would be a good relative benchmark.
I have never ever ever had to do a binary search in my life!!
Doesn't mean some else didn't have to
@@gandalfthegrey2777 Yes, true. That was not my point though. I do not think its a must to know in real life. You would know if you need it, I guess.
Yeah but data structure like binary tree are cool to learn but never implemented in my life
@@eunesshshahithakuri7047 never implemented but you use it all the time using a language api
I've implemented it once. I pulled in data from a crypto website and each coin/token started it's data on a different day. So I just picked the first day the company existed to check and then used binary search to find that first day where the data started flowing, for each token.
So far the only time I've implemented it.
Leetcode trains a certain set of skills, that are usually not used by most professionals in their specific careers, that’s why a lot of people talk about it not sticking.
Where’s the course you said you’ll mention in description!? 😅
FYI: AVL Trees are balanced.
Yes, AVL trees are balanced. An AVL (Adelson-Velsky and Landis) tree is a self-balancing binary search tree where the difference between heights of left and right subtrees cannot be more than one for all nodes. This property of AVL trees ensures that the tree remains balanced, leading to O(log n) search times, where n is the number of nodes in the tree. This makes AVL trees very efficient for lookup operations..
Source: Conversation with Bing, 18/11/2023
(1) en.wikipedia.org. en.wikipedia.org/wiki/AVL_tree.
The example shown is unbalanced.
it is though:
node 12's left subtree has height three, right has height two;
node 18's left subtree has height one, right has height zero;
node 8's left subtree has height two, the other has height one;
and node 5 is the same as node 18
It doesn't have to be weight-balanced (i.e. roughly the same number of nodes in the left and right subtrees), only height-balanced
"The best way to to crack an interview is to give as many interviews as you can" - Life Lesson.
5 things you need to know that can save your life in coding interviews:
1. Hashing and hash maps
2. Binary Search
3. Tree traversal
4. Two pointer problems
5. Basic DP (top down and bottom up)
He did not put it in the description, no one is suprised.
end the sentence with deeznuts or he'll never read it
Frontend Masters "The last algorithm you will ever need" .
Its on frontend master
@@hanhthien2948thank you
I found leetcode recently and it's a fun place that will throw out some puzzles. But it hasn't helped me build anything. I can make this one system REALLY efficient and consider how it could be optimized but... That's what you pull out after you've figured out how to build the whole system.
This ain't hentai over here. This is programming" - TheDPrimeagen
That stack overflow comment was goated !! X)))))
How are any of those leetcode exercises help with web stuff, front end, backend, queries, github, app architecture, mobile and so on?
Absolutely 😃
It makes you able to come up with your own, good solutions when the pre-existing ones doesn't deliver. Also makes you able to understand existing library code.
@@carlerikkopseng7172 which leetcode will teach you not to put queries in a loop, add indexes to dbs etc
"finite automaton of Leetcode problems" really caught me off guard lol!
While i like stuff like leetcode and codinggame and etc, i feel like most times it ends up pushing people into "small domain" mindsets. Which is fine for "find the shortest path from x to y". But at some point I'll need you to "extend large domain x with new feature z. x is already a s**tshow of a system, how would you go about implementing z without making x even more of a s**show?". Now i might need an answer that totally runs counter to "best at the small domain" but is "best at the larger domain". Can the person do that shift if he has focused too tightly on the "small domain" issues? Can he understand that many times, there are tradeoffs? Has he become so familiar with the optimal solution that now he can't see he's trying to fit a square peg into a round hole?
Loved your course on frontend masters, it helped me a lot .
For simple questions, I suppose it’s better to ask ai than searching anywhere else like stackoverflow. In 99 of 100 cases it perfectly gives you a simple example and brief explanation that’s enough to proceed further on your own. Another option is to look at docs, but sometimes it just takes a bit more time to find what you need than ask ai.
You also don't have to deal with some egomaniac being like "Simply... blah blah." and treating you like a dumbfuck for not finding a necro thread from 5 years ago on the same topic.
@@ezpk-true. Sometimes it really annoys to get extremely stupid “assumptions” about the problem on stackoverflow.
I just had an interview at a large investment banking company for an entry-level position. Throughout the interview, the interviewer kept saying, "If you don't pass this, you won't make it to the next round." Twice, he muted his microphone and turned off his video. While I was drawing on CoderPad's drawing pad, I was instructed to start coding. When I asked clarifying questions, I wasn't given any guidance, only "yes" or "no" responses. I usually visualize the problem first by drawing it, then write pseudocode, but I wasn't given the chance. I really wanted the job and was looking forward to the interview. It was disappointing, but now I just have to improve.
Stop gaslighting people. Binary search isn't easy at all. Quote from StackOverflow:
Apart from Knuth's quote that "Although the basic idea of binary search is comparatively straightforward, the details can be surprisingly tricky", there is the staggering historical fact (see TAOCP, Volume 3, section 6.2.1) that binary search was first published in 1946 but the first published binary search without bugs was in 1962.
Agree, I think best thing one can do with binsearch on interview is to state that they can be some one off errors there and use example to make sure it's fine
"Before jumping into leetcode to get good at interview questions, make sure you actually have taken a class or two on coding or finish your degree". I'd hypothesize that actually sitting there in the interview thinking about the questions does slightly more than just solving leetcode problems. Not that you wouldn't encounter similar basic tests of your understanding, but just completing coding problems is different from what many interviewers want. I don't think they want just a guy who can solve their problem, they want a guy who demonstrates their process, the how and why and is good at explaining what they're doing and why. The why. That's where you really get tripped up about your knowledge, you might be technically breezing through problems, but it might just be that you don't know what you're doing and why. The employer usually wants you to know what you're doing when you do it even if you can't solve all the problems right away. And like I believe the big O questions are popular just because it demonstrates your understanding of what the impacts of your solutions are.
I love the pen and paper point. For some reason (maybe obvious reasons) university notes and recommendations have moved from paper to virtual and I think it shows with my classmates. People don't necessarily write the notes anymore because the pdf is there, or when solving they don't start from the basics like understanding the situation and making a note on what they need to solve and what they need for solving it. They don't construct a plan (I'll admit I used to be really bad at that too and still am often too lazy, but I've reached a point where I know the subjects well enough to follow the intuition without getting tripped) or revisit theory/examples behind the topic if they're not sure how the phenomenon works. And I think it's a similar problem with the technical ability in coding interviews vs getting asked the tough questions like "have you thought about what you're doing?" I've recently started to check out a bit of design methodologies in manufacturing and trying to read the instructions through and framing the problem before attacking it. Just to not do mindless unnecessary work and get to the goal in a more coherent way. The biggest threat to that is if you can just about keep all the related concepts in your head about the problem so you won't make notes about a plan and then you forget something or get confused. And personally I'm really bad at evaluating my results before calling it done. I can often afterwards reason about them successfully to at least gauge how realistic it is, but throw it off the table as soon as I reach the answer. And that should be a priority skill.
Definitely useful advice on this video if you apply them with thought, not all are relevant to just you, but plenty of good things just for having a structured process of learning. Which is the most difficult thing to have if optional, being organised.
Where is the link for algorithm course he is talking about?
The only time I've copy pasted from SO directly and it just worked was for this problem: the percentage of overlap between rectangle A and B, given that rectangle A is always bigger, and all we have is their bounding box coordinates.
Someone happened to have the same exact question and there was a super elegant function that accepted two objects as arguments with the same name for keys.
It felt like cheating and it has never happened again lol
No link in the description :[
I wish I could get that many interviews like he mentions in the beginning, I don't even get that many so that adds extra stress to the few that I do get, which have been very very few so far.
Is this the guy that works at that Netflix startup™?
Algo courses by Prime are on frontend masters but not free as of now. Course Name: The Last Algorithms Course You'll Want (part 1 & 2)
Who has the link for the 9.5 hour algo course? I don't see it in the description
I took your advice of taking interviews at places I dont care about but it backfired and now I'm getting calls from them asking me to come back
Random indian guy: i solved 9999 leet code questions but i gonna shitcode in production. enjoy
That's just racist.
Learn data structures and programming fundamentals , do 50 leetcodes question easy and medium , get array styles and dp expirences and go for interviews , failing three interview is no different than doing thousand leetcode questions . use pen and paper , talk it loud like interview process using whiteboard . important talk through your thinking and do let know how you think .
step 2 : after solving problem in 1 hour see other people solution , how its work what unique about that etc .
aVOID BUSYNESS - DONT NEED TO record everything , need no to use notion , obsidian and extra apps for record everything and organize .
stack overflow - skill issue , know when to use apis .
can use timer to get interviewer view , and zone or can apply for leetcode contest for get in zone -- need that environment with people watching and timer .
goal : leetcode in mediumi in 15 minutes , and how good u are at solving new problem . 3 /5 in contest .
try to get good three open commit on open source . choose a technology and try it .
Only thing I ask for in interviews is to use good variable names.
I've seen people use "start" and "start_time" in the same block - often even confusing themselves.
If it's hard to get to the interview stage (i.e the job market sucks) there are discord communities and online resources for mock interviews!
I'd add to use a language that has a lot of shorthand built in so you can code faster. Like Python and their list comprehension.
Nice to be able to transpose a matrix in one line of code.
not me on Google searching "DP meaning" when he said "DP stands for Dynamic Programming YOU LOSERS!" 🤣
prime as a typical MANGA dumba is like: "NOOOOO, YA CANT CODE WITHOUT BEING ABLE TO CRACK HARD LEETCODE TASKS", and all the adequate 99% middles/seniors are like: "My code goes brrrrrr"
He has explained this multiple times. There is a secret handshake in tech that basic algorithms and data structures is what is required to get the job since it is a skillset everyone needs. If you dont want to do the secret handshake, good luck getting a dev job somewhere but it wont be at the top tech companies.
@@TJackson736 "skillset everyone needs" and "top tech companies" are hardly related to each other especially when we talk about algos and data structures. Only interns and database devs need them lol
imo if you want to actually be good you should learn the basics at the very least. you dont need to learn convex optimization or anything but leetcode pretty much covers those basics
unless you want to stick to web dev code monkeying your whole career you will need these skills at some point
tbf though i just described ~50-60% of FAANG devs. newbies ooh and aah at big tech but you really dont need to be good to work there
@@gianni50725 leetcode is so complex actually. You won't need to use algos irl in 99.99% of cases
And web dev is the hardest field for now. Nowhere else you'll face such amount of complexity and different tools
Great advices.. I have been watching ur videso before sleeping. Tomorrow I'm goin to watch this after waking up.. And thumbs up to you man..
More important, unless you're interviewing with a company where everyone speaks Hindenglish, learn proper pronunciation of the language that people speak at the workplace. Communication is important, and they might not tell you why you didn't get the job, but this could be the reason. Just knowing the words isn't enough, you have to make the correct mouth sounds.
Bro I really liked your idea about applying and practice. The greatest advices ever for gaining confidence, thanks sooo much
hey man wheres my free algorithm course???
I love the seven different people who literally just copied Neetcode’s video and thumbnail, super cool
where is the link to the free algorithms course??
"Go and interview at places you would never work at"
Sure, I mean, it takes about 1,000 applications to get one interview, but we can waste two or three interviews just to get the practice in.
Don't forget taking time off of your current job.
The entire system is designed to avoid having to hire Computer Science students. Which is actually fine. If you have a degree OR can prove proficiency through this type of interview, that helps out people who were denied a good formal education thanks to lousy K-12 schools. Everyone should have a path to catch up regardless of what zip code they were born in.
There are plenty of jobs that are less demanding that are more about data where you don't need to know binary sorting off the top of your head. But you're going to struggle with business logic if you don't have a good analytical mind.
A math degree is also very valuable in the tech industry.
True. No amount of leetcode can prep you. 4 years of experience, and I'm on my 5th company. Amount of interviews, idfk anymore.
And yes, I still get the jitters.
Where’s the link to the dsa course?
geek for geek is good for just quick revision and understanding a concept for starter
So we're focused now on developing skills for interviews instead of becoming a proficient Engineer?
yea
Because anyone can just learn the majority of the skills on the job, but you can't learn those "on the job" specific skills if you aren't good at interviews, salary negotiations, or resume/portfolio prep.
What does proficient engineer entail?
welcome to 2023.
"applying for the companies that you did not care about": worked till 2022. Now when everybody applies everywhere, you have no option anymore to interview "for free".
he definetly looks like someone who has solved 1583 leetcode questions
I feel like people should try researching pure mathmatics. everything is a whiteboard. an interview is just a collaborative puzzle solving