Tarun Gudipati

Most Asked Leetcode Problems from me in 50+ Interviews I had given. 1. Two Sum with all Approaches (https://lnkd.in/gvzUqbBi) 2. Buy and Sell Stock Sell || (https://lnkd.in/gEdS9-W6) 3. Target Sum (https://lnkd.in/gFiqGN2N) 4. LRU Cache (https://lnkd.in/gSyvaxhY) 5. Merge Intervels (https://lnkd.in/gwwqB-tN) 6. Detect Cycle (https://lnkd.in/g8XyyDsC) 7. Maximum Subarray (https://lnkd.in/gw4u3QDX) 8. Number of Islands (https://lnkd.in/gAc2UWek) 9. Path with Minimum Effort (https://lnkd.in/gTgfgmBV) 10. Trapping Rain Water (https://lnkd.in/g-AcURys)

324

3 Comments

Why did Zerodha make the shift to Golang? The answer lies in the significant advantages that Goroutines in Go offer over Java threads. Goroutines are lightweight, efficient in memory usage, and have low context switching costs, making them perfect for high-concurrency scenarios such as trading platforms. In comparison, Java threads consume about 1 MB each, while Go goroutines start with only 2 KB, enabling millions to run concurrently without draining system resources. The Go runtime scheduler effectively manages goroutines, reducing overhead and ensuring optimal CPU utilization. Moreover, Go's built-in channels streamline synchronization and communication between concurrent tasks, enhancing code safety and reducing complexity. In conclusion, Go's concurrency model and performance characteristics position it as a superior choice for real-time data streaming and high-performance applications.

38

4 Comments

India Rank 1, Global Rank 90 in #Atcoder #Beginner #352 Instead of stream, I will write my solution ideas here. I ended up solving ABCDE and G. I could not debuging WA on 2 tests for F. Here is what I did for all the problems. A: If X < Y, then we will take inbound train, start at station X and end at station Y, so station Z must be between X and Y. X<=Z<=Y must hold. If X >= Y, then we will take outbound train, start at station Y and end at station X, so station Z must be between X and Y. Y<=Z<=X must hold. B: We are given that S is subsequence of T, we should print all the index using which we can form S. Also, notice that indirectly guarantees that there will be a unique set of indexes. We can use greedy to find set of thoses indexses. This is exactly same problem B of previous codeforces round (namely round 943) C: Head of Pn is at height sum(Ai) + B_{pn} - A_{pn} So, we will have to find maximum value of B_{pn}-A_{pn), and print sum(Ai)+mx D: We can use sliding window technique here. For each i, we will maintain set of index where [i,i+k-1] exists. Then check difference of minimum and maximum value in the set. When we want to find new set for j=i+1, we erase position of p[i] and add p[j+k] E: Eaach clique gives us K*(K-1)/2 edges, and most of them are redundant, so we should ignore them. For each clique we will add K-1 edges in our graph, Ai,1 -> Ai,j with weight Ci for each 2<=j<=K_i. If graph is disconnected we should print -1, otherwise we can using any MST finding algorithm to MST of this graph. G - Lets first try a bruteforce idea. Let p_j be the probability that we have select j socks and all of them are unqiue? Then answer is 1+sum(p_j) 1<=j<=n Let tot = sim(Ai), ie total no of socks. To find p_j, let the socks we have after j selects is set c1,c2,...,cj If ck is selected at kth sock 1<=k<=j, then its probability is A_{ck}/(tot-k+1), denominator is no of socks remaining after selecting first k-1 socks. Now notice that value of numerator and denominator is same for this set irrespective of the order of selection. Numerator is prod(A_{ck)} and denominator is prod(tot-k+1) 1<k<=j So we can fix the set, find prod of all values of A_i in selected set, and then arrange them using j! ways. Let F(x) = prod(1+Ai*x) = fj*x^j Then this fj is sum of all possible value of numerators. p_j = fj*j!/(required denominator) We can use divide and conquer with NTT to find F. Its similar to build function for segment tree. Following is the pseudo code build(l,r){ if(l==r) return {1+al} mid=(l+r)/2 return convolution(build(l,mid),build(mid+1,r)) } F - I passed 60/62 tests for this problem. I could not debug corner testcase. I will write my idea in comments because I have exhaused char limit for a linkedin post.

615

23 Comments

Recent Google Interview Question - “How do you check if a graph is cyclic or not?” Now, many of us try to solve this problem using the graph algorithms like BFS/DFS, but there exists a more efficient way of solving this - DSU 🚀(Disjoint-Set Union - one of the most complex DSA topics) 🔸Start by creating a parent[] array to track all subsets. 🔸Traverse all the edges: 🔸Check the subset each node belongs to by following the parent[] array until the node and its parent are the same 🔸If two nodes are already part of the same subset, a cycle exists in the graph. Print “Cycle exists in the graph” 🔸Else, perform a union operation on those two nodes or subsets. If no cycle is found, print “Cycle doesn’t exist in the graph”. 📍Never heard of DSU? Don’t worry, here’s a Masterclass by our Codeforces Master (rated yellow) - Manas Kumar Verma Sir, to cover this advanced DSA topic which is often asked in top tech giants! ⏰ Date- 15th June, Saturday at 4 PM 🔗Secure your spot now - https://bit.ly/3KEQSgT #dsa #coding #developer #software #faang #java

1,137

33 Comments

🔹 LLD Notes GitHub Repository Nowadays, system design interview rounds have been made almost mandatory for senior roles. Although I primarily recommend reading books by Alex Xu to someone preparing for system design rounds, having a short revision material with scenario based example questions is equally important I came across this GitHub repository last week which contains compilation of tech articles, and good number of scenario questions of easy medium and hard level for LLD. I have gone through the questions, and even though the coding implementation is in Java, I feel the resource can be used for quick revision of the topics before interviews #systemdesign #interviewpreparation #technology #programming #softwaredevelopment #ig

29

2 Comments

Top 3 things that developers don’t implement in their real jobs but ask in interviews: a. Implement trie in DSA round. b. Implement Parking Lot Problem in the LLD round. c. Design a distributed system with 99.9999% uptime, infinite scalability, and zero latency in the System Design round. What do you think is the most overrated question asked in tech interviews? #techinterviews #interviewquestioninflation

84

18 Comments

In my personal experience over the last six years, working as SDE in a startup, at an established company, and a sillicon valley giant, and then running my startup, I found that engineers who are extremely good at writing optimized code and solving DSA problems are better at implementing the real stuff too and are overall better SW Engineers. This is compared to those who are just good at specific stacks like React / MERN or Django, or those who use jargon and draw diagrams in system design interviews. There is a reason why big tech companies in Silicon Valley still ask about DSA even for senior positions.

1,354

56 Comments

As per my experience with various companies I have observed four types of system design questions they asked. Category 1:- focused on how you design by keeping in mind the global presence and how you ensure scalability, availability, latency at global scale i.e. by considering presence in various geos. example: Google map, YouTube, Amazon etc. Category 2: they focus more on how well you can design data models and APIs. Focus on clean database table design and API design with proper http methods, what are the requests and responses, headers and what is the minimum possible parameter. Category 3: Just ask questions like how you design a Library like redis, scp, ssh. It involves discussion around http methods, port, socket and , IP layers and some low level detail like multithreading etc. Category 4: you release your system as a bundle of backend, frontend and database together to run a standalone server. you also provide a solution to create the cluster of servers and external database and cache. example :- JIRA and other Atlassian products. Based on the observation it would be helpful to answer and you can better craft your design. If you know any more please share in the comments. if you need help in interview preparation feel free to connect. #systemdesign #systemdesigninterview #systemdesignpreparation #softwareengineering #interviewpreparation #google #amazon #Atlassian #rubrik #faang #rippling #Microsoft

90

2 Comments

Adding more to Anshu Sharma thoughts! Why should you start your own company over working in a job? If you had started a company like NVIDIA 20 years back, the company would have grown to 3 trillion today! The value of stocks you would have as a founder would be over $ 90 billion. Assuming a 3% stake, 3% * 3 trillion. Now imagine how long would it take you to make this money if you had joined a stable company that consistently gives you 7 - 10% appraisal every year.

9

Senior developers can stay connected to the codebase, provide productive hands-on experience, maintain updated technical skills, and remain highly productive even without coding regularly. However, over time, not writing code or delivering key features can lead to a sense of moral high ground because they are not involved in meeting strict customer deadlines. On a personal note, I get little to no time to write code, but I ensure to contribute to at least one project per quarter, just for the love of it. It is one of the few job profiles that allows individual contributors to operate at such top levels. Making the most of it. Regardless, interesting read.

12

Engineers hate Mondays. That's why I'm giving away free food: 0. How Uber Tracks Billions of Trips: https://lnkd.in/da-c_aKc Author: Arpan G. 1. UML Class Diagram Explained with Examples: https://lnkd.in/dcN--x_8 Author: Ashish Pratap Singh 2. How to Approach Scaling Your Database: https://lnkd.in/degTqAxi Author: Nikki Siapno 3. How to Choose Between Batch, Micro-Batch, and Streaming when Building a Data Pipeline: https://lnkd.in/dQWkXVHq Author: Zach Wilson 4. How Deep is your DOM?: https://lnkd.in/dBcPMz8m Author: Maxi Ferreira 5. How Stripe’s Document Databases Supported 99.999% Uptime With Zero-Downtime Data Migrations: https://lnkd.in/dkpd-A4Y 6. Load Balancing Crash Course: https://lnkd.in/dqvGA65U 7. How Uber Developed a System to Manage Flaky Tests at Scale: https://lnkd.in/dcZ_ExhY Still hungry? I am packaging the other deep dives and trends in today's issue of Hungry Minds. Join 18,001+ software engineers from big tech to startups here: https://lnkd.in/dkJiiBnf What are you reading this week? #softwareengineering #coding #programming

215

51 Comments

Earlier, I talked about Uber using LedgerStore to store all the financial transactions leading to ~2 trillion indexes! 🔥😯 Check my previous posts to understand why they use hot/cold storage design to store their ever increasing ledgers. (Link in the comment) 👇 One of the first solutions they tried out was using DynamoDB. Let's understand one fact when it comes to time series data like ledgers - they exhibit a strong correlation with the current wall clock time. This can lead to clustering of write operations around a specific time range. 🕓 If we considered partitioning of data based on some time granularity, it can overload a single partition in DynamoDB, creating a "hot partition" scenario. 🔥 Hot partition leads to: 👉 Write Throttling: DynamoDB enforces throughput limits per partition. When a partition gets overloaded, write requests are throttled, causing delays and potential data loss. 👉 Performance Degradation: The overall performance of your application can suffer due to increased latency and reduced throughput. One possible solution which instantly comes to mind - Why don't we make the time granularity of a partition finer to avoid hot partitions? What does this mean? - Earlier each partition corresponded to data belonging to a time chunk of 1 minute. Let's make it finer to be 1 second. This would reduce entries per partition. 🔻 But... What if there is an increase in the traffic over time? It can lead to instability. Another side effect of this is performance degradation during the reads. One needs to perform reads via a scatter-gather which means send read requests to different partitions and perform aggregation on top of those results. Larger the number of partitions, worse would be the efficiency of this operation. 📉 Stay tuned to see how Uber solved this issue of hot partition in my next post. 😉

415

20 Comments

🚀 Exciting News! Introducing NL2MongoDB App 🧑🏼‍💻 Are you tired of grappling with complex MongoDB queries? Say hello to simplicity with the Lyzr NL2MongoDB app! 🎉 🔍 With just a few keystrokes of your natural language query, NL2MongoDB seamlessly translates it into precise MongoDB commands. No more guessing game or syntax errors—just effortless translation! ✨ Key Features: Natural Language Input: Simply type your query in plain English, and let Lyzr do the magic! Expert Translation: Our advanced AI model ensures accurate translation, mimicking human expertise. Step-by-Step Explanation: Understand every segment of the MongoDB query with detailed explanations. 🔧 Whether you're a MongoDB beginner or an expert, Lyzr NL2MongoDB streamlines your workflow, saving you time and effort. Experience the future of MongoDB querying today! For more details: https://lnkd.in/eXQv_Q7e 🔗 Learn more about Lyzr AI : Website: https://www.lyzr.ai/ Book a Demo: https://www.lyzr.ai/demos/ Try out the app: https://lnkd.in/eeaz24ZV #Lyzr #MongoDB #AI #QuerySimplified #TechInnovation #Database #NoSQL #DataManagement #DataScience #DataAnalytics #DeveloperTools #ArtificialIntelligence #MachineLearning #NaturalLanguageProcessing #TechSolution #Streamlit #AppDevelopment #TechCommunity #Innovation #ProductivityTool #Programming #SoftwareDevelopment #CodeAutomation #DigitalTransformation

5

2 Comments

The SDE-2/SDE-3 position in some companies is just like SDE-1, so don't be too surprised. Don't make them gods of tech who are SDE-2/SDE-3 in these companies. They are not gods, they are just like you. PS: Even the compensation is just like to SDE-1. #AsheeshTalks #linkedin #SDE2 #SDE3

955

35 Comments

Doing 694 Leetcodes won't get you anywhere. 20 patterns to crush your next interviews: 0. Sliding Window: Efficiently compute a target subset in a fixed-size window. 1. Matrix Traversal: Efficiently traverse 2D matrices. 2. Two Pointers: Use two pointers to process input data from two ends. 3. Fast and Slow Pointers: Detect cycles or find the middle in linked structures. 4. Merge Intervals: Work with intervals by merging overlapping ones. 5. Cyclic Sort: Sort numbers when the range is known. 6. In-place Reversal of a LinkedList: Reverse linked lists in-place for space efficiency. 7. Tree BFS: Level order traversal of trees. 8. Tree DFS: Depth-first traversal of trees for various tree-related problems. 9. Two Heaps: Use two heaps to find the median in a data stream. 10. Subsets: Generate all subsets of a given set. 11. Modified Binary Search: Customize binary search to fit problem requirements. 12. Bitwise XOR: Use XOR operations for unique element problems. 13. Top ‘K’ Elements: Use heaps to find the top K elements. 14. K-way Merge: Merge multiple sorted arrays. 15. Topological Sort: Order tasks based on dependency using graph algorithms. 16. Knapsack (Dynamic Programming): Solve optimization problems using dynamic programming. 17. Fibonacci Numbers: Simpler cases of dynamic programming (staircase...) 18. Palindrome Subsequence: Techniques to find or check palindromes with string combinations. 19. Longest Common Substring: Find the longest substring shared by two strings. What did I miss? Credit: Design Gurus #softwareengineering #interview #programming _____ Join my free software engineering newsletter: https://lnkd.in/dkJiiBnf

409

58 Comments

Leetcode Problem Of The Day : Double a Number Represented as a Linked List Given the head of a non-empty linked list representing a non-negative integer without leading zeroes. Return the head of the linked list after doubling it. Approach: 1. Reverse the Linked List: Define a method to reverse a singly linked list. Initialize three pointers: prev, curr, and next. Iterate through the list, reversing the pointers as follows: Set next to curr.next. Set curr.next to prev. Move prev to curr and curr to next. Return the prev pointer as the new head of the reversed list. 2. Double Each Node Value: Call the reverse method on the input linked list to reverse it. Initialize a carry variable to 0. Initialize a dummy node as the head of the new linked list. Iterate through the reversed list and double each node value while considering the carry: Calculate the sum of the current node value, its doubled value, and the carry. Update the carry by dividing the sum by 10. Update the sum by taking the remainder after division by 10. Create a new node with the updated sum as its value. Append the new node to the new linked list. Move to the next node in the reversed list. If there's a remaining carry after the loop, append a new node with the carry value to the new linked list. Reverse the new linked list using the reverse method. Return the head of the reversed new linked list. Time Complexity : O(n) Space Complexity : O(1) Problem Link : https://rb.gy/7g8x6s #Coding #CompetitiveProgramming #dsa #DatastructuresAndAlgorithms #Leetcode #Consistency #Programming #CodingChallenge

12

Linkedin wo platform hai jha pe Faang wali didiya ne hadd kr rakhi hai. Apne photo pe filter laga lagake course bech rahi hai. Same content sb didiyo ka exact same : 1. I failed in many interview 2. Purchase course from LossCode Academy. 3. Double filter lagake kheechi hui photo beauty plus krke. 4. Kahi bhi ye mention nahi krna ki didi LossCode se nahi pdhi hai or ye advertisement krri hai. 5. Inke course se inke filter ka koi relation nahi hai bss chipka dena hai har cheej me photo. 6. Chat gpt ka motivation 200 word ka jisko koi pdhta nahi h inke snap wale photo ko dekh ke comment kr dete hai beautifull.

142

29 Comments

How preprocessing saves tons of time for ETA calculation | Contraction Hierarchies Hey folks! Have you ever wondered how apps like Uber always seem to find the fastest way to get you from point A to point B? Today, we're going to uncover a bit of that magic by talking about something called “Contraction Hierarchies”. Imagine a Giant City Map Think about a huge city map. There are roads everywhere, connecting different places. If you wanted to find the quickest route from your home to school, you'd have to look at all these roads and figure out which ones to take. Sounds like a lot of work, right? What Are Contraction Hierarchies? Now, let’s sprinkle a little magic on this map. Contraction Hierarchies (CH) are like a super-smart way of shrinking this giant map while still keeping all the best shortcuts. Here’s how it works: 1. Simplifying the Map: First, we start by removing the least important roads. Imagine you’re taking out all the tiny alleyways and leaving only the main roads. This makes the map less cluttered. 2. Adding Shortcuts: For every road we remove, we add some "shortcuts" to make sure we can still get everywhere quickly. It's like having secret passages that only show up when you need them. 3. Building Levels: We then rank the roads. Main highways get a higher rank than small streets. This ranking helps the algorithm to decide which roads to prioritize when finding the best route. How Does Uber Use This? Uber uses something very similar to find the quickest route for drivers and passengers. Here’s a simple breakdown: - Fast Decisions: By using Contraction Hierarchies, Uber’s system can make lightning-fast decisions about which route to take. It doesn't have to look at every single road—just the important ones and the secret shortcuts. - Handling Traffic: Traffic is like a surprise obstacle in a game. Uber can quickly recalculate routes if there’s a traffic jam because it has a simplified map with lots of hidden shortcuts ready to use. Why It’s So Cool Contraction Hierarchies are like having a superhero map that always knows the best way to go. They make sure you spend less time stuck in traffic and more time enjoying your pizza. So next time you hop in an Uber and zoom through the city, you’ll know a little secret about the magic behind the scenes. It's all about making smart choices fast, just like a pro gamer! Stay curious and keep exploring! #ContractionHierarchies #uber #pathfinder #etacalculation #GPS #GoogleMaps #RoutePlanning #NodeContraction #Pathfinding #Navigation #SystemDesign #AlgorithmOptimization #ComputerScience #TechInnovation #EfficientRouting

10

 𝗔𝗿𝗲 𝗮𝗻𝗶𝗺𝗮𝘁𝗶𝗼𝗻𝘀 𝗴𝗼𝗼𝗱? This elegant, seamless animation shared in the post below is very appealing. But is it comforting when users are browsing the site and toggling between pages multiple times, especially on a site like Airbnb? Such intense animation will surely cause discomfort to motion-sensitive people. Especially for a travel app, when users are walking or moving in a vehicle if they constantly see these animations for a long duration, they may experience motion sickness. 𝗗𝗼𝗲𝘀 𝘁𝗵𝗮𝘁 𝗺𝗲𝗮𝗻 𝗮𝗻𝗶𝗺𝗮𝘁𝗶𝗼𝗻𝘀 𝗮𝗿𝗲 𝗻𝗼𝘁 𝗴𝗼𝗼𝗱? In my opinion, animations are beneficial when you are trying to convey an experience or tell a story about your product. However, if your website or app provides a large catalogue of items where users are expected to browse through many options before making a purchase, then it is not a good decision to have animations on your website. Be very particular when adding animations to your website. While animations might enhance the look and feel and make your site very appealing, they might also impact the overall interactive time on your website, which you definitely want to increase. Some supporting research says: 1. 𝗗𝗶𝗴𝗶𝘁𝗮𝗹 𝗘𝘆𝗲 𝗦𝘁𝗿𝗮𝗶𝗻 (𝗗𝗘𝗦): Prolonged use of digital screens, including viewing animations, can lead to digital eye strain (DES), characterized by symptoms such as dryness, irritation, blurred vision, headaches, and difficulty focusing. This is especially prevalent among individuals who use computers and smartphones extensively. 2. 𝗠𝗼𝘁𝗶𝗼𝗻 𝗦𝗲𝗻𝘀𝗶𝘁𝗶𝘃𝗶𝘁𝘆: Users with motion sensitivity can experience discomfort and even nausea from intense animations. This is particularly problematic for applications used in dynamic environments, such as travel apps, where users might be on the move and more prone to motion sickness. 3. 𝗜𝗺𝗽𝗮𝗰𝘁 𝗼𝗻 𝗣𝗿𝗼𝗱𝘂𝗰𝘁𝗶𝘃𝗶𝘁𝘆: Digital eye strain affects comfort and reduces productivity and accuracy in tasks performed on digital devices. Continuous exposure to animations and other dynamic elements can exacerbate these issues. 4. 𝗗𝗲𝘀𝗶𝗴𝗻 𝗥𝗲𝗰𝗼𝗺𝗺𝗲𝗻𝗱𝗮𝘁𝗶𝗼𝗻𝘀: To mitigate these effects, it is advised to use animations sparingly and strategically. Simple and purposeful animations that enhance user experience without overwhelming them are recommended. Additionally, providing users with options to reduce or disable animations can help cater to those who are sensitive to motion. #UserExperience #WebDesign #UXDesign #MotionSensitivity #WebDevelopment #Accessibility #TechHealth #Productivity #TechTips #WebAccessibility #DesignThinking #UserComfort #WebAnimation

17