Twitch | On Site | Full Stack Engineer [Reject]

twitch logo
twitch
Full Stack EngineerRejected
December 12, 20201 reads

Summary

I recently went through a multi-stage interview process for a Full Stack Engineer role at Twitch, which included an online assessment, a phone screen, and a virtual onsite covering OOP, data structures & algorithms, system design, and behavioral aspects. Unfortunately, I was ultimately rejected after a week-long wait for the feedback.

Full Experience

I recently went through the entire interview process with Twitch for a Full Stack Engineer position.

My journey began with an Online Assessment. This round was primarily focused on Object-Oriented Programming, and the task was to implement a Battleship game. It took me some time, but I managed to complete it successfully.

Following that, I had a 60-minute Phone/Google Meet interview. This round involved discussing my background, answering a few basic technical questions, and solving a coding problem related to Data Structures & Algorithms. I found this part relatively easy.

The next stage was a comprehensive 4-hour virtual onsite interview, which consisted of four back-to-back sessions. Most of these were technical, with one dedicated behavioral interview.

  • The First interview presented a unique coding challenge. I was given a list of items, each with a value and boolean flags indicating start and end. My task was to implement a function that would identify and process 'full valid sets' of items. A valid set meant finding all items sequentially between a start: true item and an end: true item. For instance, in a list like { val: 0, start: false, end: false }, { val: 1, start: true, end: false }, { val: 3, start: false, end: true }, { val: 2, start: false, end: false }, { val: 7, start: false, end: false }, the valid set would only include items 1, 2, and 3. I approached this problem using a priority queue.
  • The Second Interview was a basic Depth-First Search (DFS) question. The problem was similar in concept to the well-known 'Number of Islands' problem.
  • The Third Interview was purely behavioral. I was asked standard questions such as 'Tell me about a time you helped a peer get unstuck' and 'Tell me about a time you solved something challenging.'
  • For the Fourth Interview, the focus shifted to System Design. I was tasked with designing a full-stack component for a video and a 'like' button. This involved explaining the architectural choices and providing some code snippets to illustrate my approach.

After completing the onsite, I was initially told I would receive feedback the very next day. However, I had to proactively reach out a week later to hear back. Unfortunately, I received a rejection, but I consider it a valuable learning experience and appreciated the process.

Interview Questions (6)

Q1
Battleship Game Implementation (OOP)
Data Structures & Algorithms

The online assessment required me to implement a Battleship game program. The primary focus was on demonstrating strong object-oriented programming (OOP) principles and design patterns.

Q2
Extract Valid Item Sets from Stream with Start/End Flags
Data Structures & Algorithms

Given a list of items, where each item has a value, a start boolean flag, and an end boolean flag, I needed to implement a function. This function's goal was to identify and collect 'full valid sets' of items. A full valid set is defined as all items encountered sequentially between an item with start: true and an item with end: true, inclusive. For example, given the stream: { val: 0, start: false, end: false }, { val: 1, start: true, end: false }, { val: 3, start: false, end: true }, { val: 2, start: false, end: false }, { val: 7, start: false, end: false }, the function should identify items 1, 2, and 3 as a valid set (assuming 2 came after 1 and before 3 logically for the set). Once a full valid set is accumulated, a predefined callback function should be invoked with this set.

Q3
DFS - Similar to Number of Islands
Data Structures & Algorithms

I was given a coding problem that involved Depth-First Search (DFS). The conceptual nature of the problem was similar to the classic 'Number of Islands' LeetCode problem, where one typically traverses a 2D grid or graph to find and count connected components.

Q4
Behavioral: Helping a Peer Get Unstuck
Behavioral

The interviewer asked me to describe a specific situation where I successfully helped a colleague or peer overcome a technical challenge or get unstuck on a project.

Q5
Behavioral: Solving a Challenging Problem
Behavioral

I was asked to recall and elaborate on a challenging problem I encountered in my professional experience and detail the steps I took to solve it.

Q6
System Design: Full Stack Video and Like Button Component
System Design

The task was to design a comprehensive full-stack system for a video viewing platform, specifically focusing on the video playback functionality and an associated 'like' button. This required outlining the architecture, detailing API endpoints, database schemas, and discussing front-end and back-end interactions. I also had to demonstrate key logic with some code.

Discussion (0)

Share your thoughts and ask questions

Join the Discussion

Sign in with Google to share your thoughts and ask questions

No comments yet

Be the first to share your thoughts and start the discussion!