You are going on a camping trip, but before you leave you need to buy groceries. To optimize your time spent in the store, instead of buying the items from your shopping list in order, you plan to buy everything you need from one department before moving to the next.

Given an unsorted list of products with their departments and a shopping list, return the time saved in terms of the number of department visits eliminated.

Example:

products = [
    ["Cheese",          "Dairy"],
    ["Carrots",         "Produce"],
    ["Potatoes",        "Produce"],
    ["Canned Tuna",     "Pantry"],
    ["Romaine Lettuce", "Produce"],
    ["Chocolate Milk",  "Dairy"],
    ["Flour",           "Pantry"],
    ["Iceberg Lettuce", "Produce"],
    ["Coffee",          "Pantry"],
    ["Pasta",           "Pantry"],
    ["Milk",            "Dairy"],
    ["Blueberries",     "Produce"],
    ["Pasta Sauce",     "Pantry"]
]

list1 = ["Blueberries", "Milk", "Coffee", "Flour", "Cheese", "Carrots"]

For example, buying the items from list1 in order would take 5 department visits, whereas your method would lead to only visiting 3 departments, a difference of 2 departments.

Old:

Produce(Blueberries)->Dairy(Milk)->Pantry(Coffee/Flour)->Dairy(Cheese)->Produce(Carrots) = 5 department visits

New:

Produce(Blueberries/Carrots)->Pantry(Coffee/Flour)->Dairy(Milk/Cheese) = 3 department visits

list2 = ["Blueberries", "Carrots", "Coffee", "Milk", "Flour", "Cheese"] => 2
list3 = ["Blueberries", "Carrots", "Romaine Lettuce", "Iceberg Lettuce"] => 0
list4 = ["Milk", "Flour", "Chocolate Milk", "Pasta Sauce"] => 2
list5 = ["Cheese", "Potatoes", "Blueberries", "Canned Tuna"] => 0

All Test Cases:

shopping(products, list1) => 2
shopping(products, list2) => 2
shopping(products, list3) => 0
shopping(products, list4) => 2
shopping(products, list5) => 0

Discuss the approach for implementing an LRU (Least Recently Used) cache. Coding was not required, only an architectural discussion.

Design a collaborative editing system similar to Google Docs. The discussion should focus on conflict handling mechanisms and the use of WebSockets for real-time updates.

Design a file storage and searching system. Files should be searchable by file name or by keywords within their content. The proposed solution should leverage object storage for files and Elasticsearch for content-based keyword searching.

Design a web scraping system capable of scraping a given list of URLs and extracting all image URLs from them. The system should expose the following APIs:

1. POST /jobs -> {urls: ["abc.com", "amazon.com"]}  -> response ({job_id: 1234})
2. GET /jobs/{job_id}/status -> {"completed": "1", "inprogress": "1"}
3. GET /jobs/{job_id}/results -> {"https://abc.com": ["abc.com/img1.jpg"],"https://amazon.com": ["amazon.com/img1.jpg"]}

You are the manager of a cinema hall. Design a system to determine if a new movie can be added to an existing schedule without removing any current movies. The cinema operates from 10 AM to 11 PM on a single day. Movie start and end times are represented as minutes from midnight. Implement a function boolean canSchedule(Movie movie, MovieSchedule schedule); that takes a Movie (with its duration) and the current MovieSchedule as input, and determines if the movie can be scheduled. Example data provided:

{
  "movies": [
    {
      "title": "Lord Of The Rings",
      "durationInMinutes": 120
    },
    {
      "title": "Back To The Future",
      "durationInMinutes": 90
    }
  ],
  "screenings": [
    {
      "title": "Lord Of The Rings",
      "startTime": 660
    },
    {
      "title": "Lord Of The Rings",
      "startTime": 840
    },
    {
      "title": "Back To The Future",
      "startTime": 1020
    },
    {
      "title": "Lord Of The Rings",
      "startTime": 1200
    }
  ]
}

Given an employee directory structure composed of groups and departments, find the closest common parent group for a target set of employees. The problem is analogous to finding the Lowest Common Ancestor (LCA) in a tree structure. For example, in a given diagram (implicitly a hierarchy), the closest group for employees Alice and Lisa would be 'Engg'.

I was asked to implement a Snake Game with a slight variation. Instead of increasing in size every time it eats food, the snake only increases its size after eating every 'n'th food item. The focus was on code organization, creating appropriate objects, functions, and abstractions to ensure a modular, extensible, and clean implementation. The expected return object was a snakeGame object which exposes a nextMove(dirn) function, where the user inputs the next direction for the snake to move. I needed to internally manage the state of the board, the snake, and food items.

Given an n-ary tree structure, I needed to find the Lowest Common Ancestor (LCA) for any given pair of leaf nodes. The interviewer emphasized time complexity, clear explanations, and continuous communication of my thought process.

I was asked to design a YouTube-like homepage for a software like Confluence. The goal was to display the most "engaged with" documents across the organization. Additionally, the system should not show documents that I, as the user, have already written or read, similar to YouTube's recommendation system. It was mostly a standard leaderboard question.

This round revolved around Atlassian's principles. I was expected to provide examples showcasing how I embody their guiding values, such as 'be the change you seek' by describing a time I initiated change in my company.

This round was similar to the values round but with a much stronger focus on my previous experience. The interviewer delved deeply into my past projects, questioning how I demonstrated Atlassian values within those projects. It was important to show thoroughness regarding my past work.