Google 2026 New Grad R1 and NYC Onsite: Median, Three Sum, and Dynamic Islands

Role: Software Engineer, New Grad
Result: Rejected after the onsite
Report time: August 2026
Onsite location: New York City
Application: Applied online

Timeline

1. Applied in early November 2025.
2. Received the OA two days later and passed it the same evening.
3. Completed R1 coding and behavioral interviews on the same day.
4. Learned that R1 had passed about one week later.
5. Completed two onsite coding rounds in New York in mid-December.
6. Completed a visa check in late December.
7. Received the rejection in mid-January.

R1 coding: streaming median

Implement the data-stream median problem using two heaps. After coding, the candidate completed a dry run and analyzed complexity.

Follow-ups asked:

1. What changes if the value range is very small?
2. What changes if the query asks for the kth-smallest value instead of the median?

The candidate discussed an approach and pseudocode for the bounded-range case. Time expired while discussing the kth-smallest version. A comment later asked whether k was fixed or supplied with each query, but the candidate did not answer.

Practice: Streaming Median and Kth Smallest

R1 behavioral

The interviewer did not ask about the resume. Questions included handling changing priorities and describing the role the candidate usually plays on a team.

Onsite coding 1

After about 15 minutes of resume and technical-project discussion, the interviewer asked Two Sum followed by Three Sum. The candidate completed both implementations and the complexity analysis.

Practice: Find Two-Sum or Three-Sum Index Combinations

Onsite coding 2

The base problem was Number of Islands, which the candidate solved with BFS.

The follow-up added land cells over time and asked for the changing island count. The candidate first tried to modify the BFS solution. The interviewer asked for a Union-Find solution, but the candidate did not finish before time expired.

Practice: Count Islands After Land Additions

Travel note

Google could book the flight and hotel directly. The candidate kept receipts for meals and rides, uploaded them to the reimbursement site, and reported that reimbursement was straightforward.