Google PhD New Grad Loop: Growing Islands, JSON and Table Parsing, and Trie Search

Role: Software Engineer, PhD New Grad
Result: Passed the loop and entered Team Match
Report time: June 2025

Timeline

• Applied around Christmas 2024
• Completed Google hiring assessment in late January
• Phone interview in late February
• Three coding rounds plus one behavioral round in late April or early May
• Entered Team Match in June

Phone screen: largest island and dynamic additions

Find the largest island in a grid. Then support adding land over time and merging islands as they grow.

Practice: Maximum Area of an Island

Follow-up practice: Count Islands After Land Additions

Onsite coding 1: JSON calculation

Parse a JSON file, locate specified information, and perform multiplication and addition. The exact schema, lookup path, and output were not preserved.

Onsite coding 2: nested table parsing

Parse a text table, then support tables nested inside tables. The source does not preserve the grammar or required output.

Onsite coding 3: prefix tree

Build a trie to find matching words. The exact query operation—prefix enumeration, exact lookup, autocomplete, or another variant—was not preserved.

Behavioral interview

Questions were described as standard, including how the candidate handles a coworker they do not get along with.

The candidate had an engineering PhD background spanning numerical simulation, high-performance and parallel computing, and ML acceleration, plus an ongoing Google LLM internship. They entered Team Match but reported limited ML-role availability in the eastern United States.