GOOGLE SWE L3 — FIRST ROUND TECHNICAL INTERVIEW

The Interview

On February 9, 2026 I completed my first round technical interview for the Software Engineer L3 position at Google. The round consisted of three parts: a coding challenge, a behavioral interview, and a Googleyness & Leadership assessment.

Coding Challenge

The coding problem I received was very similar to LeetCode #394 — Decode String.

Problem

Given an encoded string, return its decoded string. The encoding rule is k[encoded_string], where the encoded_string inside the square brackets is repeated exactly k times. Brackets can be nested.

Input:  "3[a]2[bc]"    → Output: "aaabcbc"
Input:  "3[a2[c]]"     → Output: "accaccacc"
Input:  "2[abc]3[cd]ef" → Output: "abcabccdcdcdef"

Optimal Solution — Two-Stack Approach

The most optimal solution uses two stacks: one for multipliers and one for accumulated strings. Single pass through the input, O(n) time.

def decode_string(s: str) -> str:
    count_stack = []
    string_stack = []
    current = ""
    k = 0

    for ch in s:
        if ch.isdigit():
            k = k * 10 + int(ch)
        elif ch == "[":
            count_stack.append(k)
            string_stack.append(current)
            current = ""
            k = 0
        elif ch == "]":
            prev = string_stack.pop()
            repeat = count_stack.pop()
            current = prev + current * repeat
        else:
            current += ch

    return current

How it works:

• Digits build up the multiplier k (handles multi-digit numbers like 12[a])
• On [ — push current state (multiplier + accumulated string) onto stacks, reset for the nested section
• On ] — pop the previous string and multiplier, combine: previous + current * repeat
• Letters simply append to the current string

Complexity: O(n) time, O(d) space where d is the maximum nesting depth of brackets.

Why This Approach Wins

A recursive solution also works but the iterative two-stack approach avoids call stack overhead and handles arbitrarily deep nesting without risk of stack overflow. It processes each character exactly once.

Behavioral

Standard STAR-format questions about past projects, teamwork, conflict resolution, and how I handle ambiguity. I drew heavily on my experience at Airbus Defence and Space — cross-functional collaboration on production workflows and the 30% assembly time reduction project.

Googleyness & Leadership

This round assessed cultural fit — intellectual humility, comfort with ambiguity, bias toward action, and collaborative instincts. Questions about navigating disagreements, supporting teammates, and taking initiative without being asked.

Takeaways

• LeetCode #394 is a classic stack problem — recognizing the pattern immediately saves time
• The two-stack approach is clean and interviewer-friendly: easy to explain, easy to trace through examples
• Behavioral prep matters as much as coding prep at Google — they evaluate the whole engineer