I was given a problem involving DNA fragments represented by a `Sequence` object with `startId`, `endId`, and `payload` fields. The goal was to implement a function `String shotgunSequence(List<Sequence> sequences)` that concatenates the payloads in the correc...
Showing 1-10 of 39 forum items
The assessment was on HackerRank and lasted 75 minutes. There were two questions: 1) Find the shortest subarray size with at least K distinct characters. 2) Given an array of n integer prices in non-decreasing order, and q queries to process. For each query, I...
I was asked to parse a CSV file into a list of row objects (dictionaries), where column headers are keys. Values enclosed in double quotes should be treated as strings, while unquoted numeric values should be parsed as integers. I had to implement query functi...
During my onsite interview, I had a coding round. I was asked to sort a DNA sequence by its starting and ending sequences. Given a list of tuples like: ``` [ (start = "bbb", end = "ccc", payload="1"), (start = "aaa", end = "bbb", payload="2"), (start = "ccc",...
I was asked about top-p, top-k, greedy search, and beam search. Then, I had to implement greedy search and beam search in Python. Beam search was a bit tricky, and there are many areas for potential improvement regarding time and space complexity. I was given...
The online assessment consisted of two coding questions to be completed in 90 minutes on HackerRank. The browser automatically entered fullscreen, with HackerRank tracking tab switches. One question was a variation of LeetCode 2389. My initial solution was not...
After a brief introduction of the business context, I was asked a series of fundamental ML/DL questions: * The differences between Layer Normalization and Batch Normalization. * Characteristics of different regularization terms. * Approaches to CTR prediction....
I had three rounds of interviews, each lasting 45 minutes and covering coding and behavioral questions. I was expected to write main methods to run my code and provide complete solutions for all follow-up questions. For behavioral questions, I was asked about...
The interviewer was the hiring manager. I had to implement K-Means clustering, but it didn't need to be runnable. The interviewer just needed to see that my implementation was okay. I could use any framework I wanted (Python/NumPy/PyTorch).
Round 1 consisted of a medium-difficulty problem from the Leetcode 150, followed by a behavioral question: How do you make decisions if you have multiple solutions? In Round 2, I was asked about debugging, error handling, and how to approach pull request refac...
