RoundHound
?Sign inAccess
Prepare
HomeHomeTargetsTargetsPractice historyHistoryProgressProgress
Get full access?Sign in
Targets/Google

Software Engineer

Google · Mid-level

Research from Sep 65 rounds, none tried
How did it actually go?

Recommended next: your first attempt

Do the Technical coding screen round.

A complete first attempt shows where you stand before any narrow practice starts.

Set up this round →Rehearse the full day
Length45 min · Mid-level

The loop

5 rounds · 3 h 45 min

Pick a round to see what it asks of you and where you stand on it.

Technical coding screen

Establish independent problem solving: turn a verbal specification into correct code and explain why it works.

Coding · 45 minSet up this round →
Not tried yetNo attempt yet.

What strong sounds like

  • Clarify the input, output and one ambiguous edge case before selecting a structure.
  • Produce complete code whose state and termination match the stated algorithm.
  • Explain the invariant and time/space costs, then manually trace a boundary case.
  • Identify what changes under the follow-up and preserve the unaffected logic.

Expect to be pushed on

What breaks with duplicates or disconnected input?
How does the solution change when the input becomes a stream?

Shapes of problem you may get

  • A traversal or interval problem with a changed input constraint

How to prepare

  1. 01
    Rehearse specification and invariants

    Select two unseen medium problems. For each, write the input, output, one ambiguity, the chosen state, and the invariant before coding. Briefly explain the invariant aloud once.

    The screen tests independent problem solving and explicit correctness reasoning. Google reports also show graph questions with changed constraints, so a stable explanation matters more than recalling one solution.

    Company researchGoogle L4: phone screen and onsite, April 2025Google L4 interview experience, May 2025Google L4 interview experience, August 2026

A first attempt takes the full 45 minutes.

What this loop tests

Researched 27 days ago

Prepare for a coding screen, three algorithmic interviews, and a Googleyness and leadership conversation. L4 coding rewards turning an unfamiliar problem into a precise model, implementing it clearly, and adapting to follow-ups. Prioritize graphs, intervals, trees, and stateful data structures; retain breadth in recursion and mathematical reasoning. Practice complete solutions with complexity analysis and manual traces, alongside concrete collaboration and ambiguity stories.

They reward

  • Precise problem models
  • Complete, readable implementations
  • Adaptation under follow-ups
  • Collaborative judgment with visible learning

What to avoid

  • Do not recite a remembered solution without checking constraints.
  • Do not leave the final complexity implicit.
  • Do not mark a state visited before defining what makes it unique.
  • Do not assume every graph is a tree.
  • Do not force dynamic programming onto a simpler problem.

How sure we are

Worth redoing if you hear something from the recruiter that contradicts this.

Attempts

Every session saved against this target, newest first. Standings live on the round cards.

All attempts in History →

Technical coding screen, Coding: trees and graph state, Coding: sequences and reasoning, Coding: stateful data structures, and Googleyness and leadership have no attempts yet.

  • 02
    Plan a verification checkpoint

    Create a short end-of-solution checklist: compile or syntax check, boundary trace, invariant check, complexity, and follow-up impact. Use it on one selected problem without completing a timed round.

    The supplied bar requires complete code, a boundary trace, and explicit costs. A fixed checkpoint protects verification time in a 45-minute conversation.

    Suggested approach

  • 03
    Prepare for constraint changes

    For each selected problem, write one likely follow-up that changes input size, edge conditions, or state. Mark which code and complexity claims remain valid and which must change.

    Reported Google L4 loops include harder constraints after an initial solution. This preparation targets adaptation without encouraging memorized answers.

    Company researchGoogle L4: phone screen and onsite, April 2025Google L4 interview experience, May 2025Google L4 interview experience, August 2026

    • Then try the round

    You have two problem outlines, a usable verification checklist, and one written follow-up analysis. You can then attempt the first practice round.

    Resources to focus on
    • Interviewing at Google: Best practices, advice, and tips

      Review the Google Careers interview preparation guidance and recruiter-provided interview information.

      Use the company guidance to confirm practical details without treating it as evidence for a specific coding topic.

    • Paths in graphs

      Read “4.2 Breadth-first search” and “Correctness and efficiency.”

      Extract a concise invariant, state definition, and complexity explanation for traversal problems.

    Do

    • State the invariant before coding.
    • Reserve time to trace your own implementation.

    Avoid

    • Do not recite a remembered solution without checking constraints.
    • Do not leave the final complexity implicit.
    Confirmed · 8
    • L4 candidates describe a technical screen followed by three coding interviews and a Googleyness conversation.
    • Graph traversal appears in both screening and onsite reports, with changes to constraints after the initial solution.
    • A May 2025 L4 candidate describes class design with additional operations and harder data-structure follow-ups.
    • An August 2026 L4 report includes trie search, array/binary search, and a modified Dijkstra problem.
    Likely · 2
    • Preparation should prioritize adaptable algorithmic reasoning over memorizing the most frequently tagged questions.
    • A strong L4 practice bar is independent implementation, explicit complexity, deliberate verification, and a coherent response to a changed constraint.

    Worth asking them

    Questions that close the gaps above.

    How does this team distinguish strong L4 ownership from L5 scope?
    Which user or reliability constraint drives the team’s hardest decisions?
    How do engineers influence work beyond their immediate project?

    Research status

    Collected Sep 6. If the recruiter tells you something this brief contradicts, correct it after the interview and the brief updates.

    Correct this brief