First round OA: LC coding + AI frontend/backend
Sliding window / hash map problem, data scale up to 10^5, so it had to be an O(N) solution.
Description:
A class has many students, and each student has exactly one specific talent. There are talentsCount total types of talent, numbered from 1 to talentsCount. Now we need to form teams for a competition, and each team must include every type of talent (i.e., at least one person for each of 1 through talentsCount).
The team has to be made up of consecutive students from the array. For every possible starting position (left endpoint) in the array, you need to find the length of the shortest contiguous subarray starting there that covers all the talents. If it's impossible to collect all the talents starting from some position, return -1 for that position.
Example:
Input:
talentsCount = 3
talent = [1, 2, 3, 2, 1]
Output: [3, 4, 3, -1, -1]
Explanation:
Starting from index 0 (talent 1): you need to see [1, 2, 3] to collect all 3 talents, shortest length 3.
Starting from index 1 (talent 2): you need to see [2, 3, 2, 1] to collect all talents, shortest length 4.
Starting from index 2 (talent 3): you need to see [3, 2, 1] to collect all, shortest length 3.
Starting from index 3 and index 4: not enough people left to collect all talents, so return -1.
Technical interview:
They asked a few frontend questions about React, and also how I use AI in my work.
There was a scenario question: during peak shopping season, tons of users flood in, the page loads slowly, and there are 404s. Horizontal/vertical scaling — add more server instances during peak traffic.
Two coding questions.
One was backtracking/DP: three parameters — max weekly hours (≤56), max daily hours (≤8), and a 7-character pattern (each character is a digit 0-8 or ?, where ? means flexible). You need to return all valid scheduling plans: fill in the ?s with digits so that each day is ≤ maxDaily and the total equals the weekly hours. Kind of like LeetCode 216, but felt harder, and I didn't finish it. The interviewer did guide me along.
Second one was merging two strings, LC 1768. Ran out of time, but I glanced at it and it looked like an easy problem.
I probably failed — the interviewer asked whether I could code in JavaScript... I did it in Python.
Discussion
Loading comments…