Google SWE II (L3) Interview Experience - India | 2026
Summary
I passed the Google Hiring Assessment, completed the DSA and Googliness phone screens, and went through two onsite rounds that focused on data‑structure and algorithm problems.
Full Experience
Google Hiring Assessment (GHA)
- The GHA had multiple situational questions around work style, decision-making under pressure, and mindset at work.
- I went through a few YouTube videos beforehand to better understand the type of questions and how to approach them.
- I found it helpful to keep my answers consistent throughout the assessment and to think in terms of collaboration, ownership, impact, problem-solving, and thoughtful decision-making.
- For workplace conflicts or disagreements, I focused on collaboration and finding a solution rather than choosing an approach that works against the team.
- For situations involving unethical behaviour or violations of workplace policies, I answered based on the relevant company policies and processes, regardless of whether the person involved was a peer, manager, or someone else.
- Some preparation videos suggested avoiding neutral answers when I could reasonably take a clear position. I therefore used strongly agree/disagree when the scenario clearly supported it and chose a neutral option only when I genuinely could not strongly agree or disagree.
Phone Screen - DSA
Longest Increasing Subsequence with Difference 1
Given an integer array, find the length of the longest increasing subsequence where the difference between consecutive elements is exactly 1.
- Discussed an O(n²) DP approach.
- Optimized it to O(n) using a hash map and coded the optimized solution.
Follow-up: Difference in a Range Find the longest increasing subsequence where the difference between consecutive elements can be from 1 to d.
- Discussed an O(n²) DP approach.
- Also discussed an O(n*d) hash‑map‑based approach when d is small.
Phone Screen — Googliness
- Questions around conflict management
- Handling missed deadlines
- Handling unclear goals
- How to deal with someone taking credit for my work multiple times
- How I use AI in my daily work
- What I would want at work that doesn't exist today, including ways to improve work/productivity and workplace perks/experience
- One thing I want to achieve in the near future in my career/work
Multiple cross‑questions and follow‑ups on my answers
Onsite 1 - DSA
Minimum Security Level to Reach a Node
Given a directed graph where every edge has a security level, find the minimum value of S needed to travel from u to v. An edge can be traversed only if its security level is less than S.
- I clarified cases such as negative security levels and cycles.
- The interviewer then asked how the approach would change for each of these different scenarios.
- He focused heavily on the reasoning behind the choice of data structure and approach, and asked me to explain each case with examples.
- We discussed queue vs. priority queue, including how the choice affects the number of operations and time complexity.
- About 30‑35 minutes were spent discussing different scenarios and approaches, leaving relatively little time for coding.
Onsite 2 - DSA
Minimum Cost to Remove All Leaf Nodes
Given a binary tree with edge weights, remove all leaf nodes from the root. A cut can be made anywhere in the tree, removing the corresponding leaf/subtree. The cost of removing all the leaf nodes is the sum of the weights of the edges being cut. Find the minimum total cost.
- I clarified that the edge weights were positive.
- Solved it using recursion in
O(n). - Initially, I represented the edge weights separately using a vector – for example, storing the edge weight for a node at the corresponding index.
- The interviewer asked if there was a better way to represent the edge weights, so I suggested storing the weight directly in the node structure and he asked me to code the required
Nodestruct/class.
Follow-up 1: N‑ary Tree
What if a node can have more than two children?
- Extended the solution to support multiple children.
- Used a vector of pair to store child nodes and their corresponding edge weights.
- Also defined the required node structure.
Follow-up 2: Negative Edge Weights
What if edge weights can be negative?
- No coding required.
- Discussed taking all negative‑weight edges since cuts are unrestricted and negative edges should always be included.
- Then, if any leaves remain, remove the remaining required positive‑weight edges using the minimum additional cost.
Timeline
July 1: Gave the Google Hiring Assessment (GHA).
July 2: Received an email confirming that I had passed the GHA.
July 3: Recruiter called to schedule the phone screens - one DSA round and one Googliness round.
July 15: DSA phone screen. I chose a slot after 4 PM, so the interviewer was based in Europe.
July 16: Googliness phone screen. Again chose a slot after 4 PM, so the interviewer was based in Europe.
July 20: Recruiter called with the results, but we couldn't connect due to a technical issue.
July 23: Recruiter informed me that I was moving to the onsite rounds. She mentioned that it would take some time to initiate the onsite process.
August 4: Recruiter was changed.
August 17: New recruiter contacted me for availability for the onsite rounds.
August 31: Both onsite rounds were initially scheduled for this date. Only Onsite 1 happened; Onsite 2 was rescheduled to September 8 because the interviewer was unavailable.
September 8: Onsite 2 was rescheduled again to September 17 due to interviewer unavailability.
September 17: Onsite 2 took place.
Interview Questions (6)
Longest Increasing Subsequence with Difference 1
Given an integer array, find the length of the longest increasing subsequence where the difference between consecutive elements is exactly 1.
Longest Increasing Subsequence with Difference up to d
Find the longest increasing subsequence where the difference between consecutive elements can be any integer from 1 to d.
Minimum Security Level to Reach a Node
Given a directed graph where each edge has a security level, find the minimum value S needed to travel from node u to node v. An edge can be traversed only if its security level is less than S.
Minimum Cost to Remove All Leaf Nodes (Binary Tree)
Given a binary tree with edge weights, cut edges to remove all leaf nodes. The cost is the sum of the weights of the cut edges. Find the minimum total cost.
Minimum Cost to Remove All Leaf Nodes – N‑ary Tree Extension
Extend the previous problem to a tree where a node may have more than two children.
Minimum Cost to Remove All Leaf Nodes – Negative Edge Weights
Consider the same leaf‑removal problem where edge weights can be negative.
Preparation Tips
I watched several YouTube videos that explained the Google Hiring Assessment format and the type of behavioral questions asked. For the DSA rounds, I practiced dynamic‑programming problems like Longest Increasing Subsequence and studied graph traversal techniques, especially variations involving edge constraints. I also reviewed tree problems that required custom node structures and thought about how to handle extensions such as N‑ary trees and negative edge weights.