← Interview experiences
Rejected

Software Engineer (L4) at Google

Difficulty
Process took
4-5 Weeks
Rounds
5
Format
Remote
Applied via
Referral

How it went

  • This experience humbled me. I realized that solving problems quickly isn’t enough, you must truly understand them, reflect on failures, and grow from them.
  • Don't ignore fundamentals. In hindsight, rushing through DSA during college and not revisiting those foundations cost me this opportunity.

What they would tell you

  • This experience humbled me. I realized that solving problems quickly isn’t enough, you must truly understand them, reflect on failures, and grow from them.
  • Don't ignore fundamentals. In hindsight, rushing through DSA during college and not revisiting those foundations cost me this opportunity.

How to prepare

  • You will rarely be asked about textbook problems. Each question was a twist on known patterns. Understanding the “why” behind algorithms is more useful than memorizing them.

Round by round

  1. 1

    Phone Screen60 min

    This round was conducted by an engineer from Google Berlin. The question, while appearing simple on the surface, required thoughtful arrangement logic Question Asked: Given an m x n matrix and a list of items with their quantities, arrange them such that similar items are placed adjacent, either horizontally or vertically. Initially, it seemed like a graph problem, but it was more about smart distribution and spatial placement strategy. I broke down the problem, discussed edge cases, and coded a working solution with time and space analysis.

  2. 2

    Technical Interview 160 min

    This interviewer was from Google India. Question Asked: In a cloud storage system, you can create disks and take snapshots. A snapshot can only be deleted after all disks derived from it (and their descendants) are deleted.Given disk-snapshot relations, return a valid deletion sequence. After some discussion and clarification, I recognized it as a DAG (Directed Acyclic Graph) problem. I applied topological sorting, discussed complexity, and wrote clean code.

  3. 3

    Technical Interview 260 min

    Question Asked: Given a list of 2D boxes with height, width, and depth, determine the maximum number of boxes that can be nested inside each other. A classic DP problem often likened to Longest Increasing Subsequence in 2D. I could only come up with the brute-force solution and explained that, but I couldn’t progress toward the optimized version in time.

  4. 4

    Technical Interview 360 min

    Question Asked: Given a neighborhood of blocks, each containing colored houses (numbers), rearrange them so that each block contains uniquely colored houses sorted in increasing order. I tried a frequency + bucket sort strategy, and had a near-working approach. In the final minutes, I realized a binary search could help optimize placement, but I ran out of time before I could refactor and implement it.

  5. 5

    Googliness & Leadership Round60 min

    It focused on:

    • My background and past projects
    • Times I showed leadership, adaptability, and learning
    • Collaboration and ownership stories The interviewer was warm and curious. It felt more like a conversation than an assessment.

What came up

MediumRemote2 3 years

A candidate-reported account, lightly edited. Interview processes change by team and date.

Sources