Google PhD New Grad SWE Onsite: Calendar Gaps, Data Chunks, ML Basics, and BQ

Role: Software Engineer, PhD New Grad — ML
Result: Entered Team Match
Report time: August 2025

Coding 1: calendar scheduling with do-not-disturb time

Schedule a meeting while avoiding do-not-disturb intervals and return all non-overlapping available periods.

The source does not define the calendar horizon, meeting duration, whether output intervals must be maximal, or endpoint semantics. Those rules are required for a deterministic practice problem, so no exact link is asserted.

Coding 2: minimum data chunks

Given a fixed amount of data and a maximum packet size, split the data into the minimum number of packets.

This is the zero-header mode of FastPrep's chunking problem.

Practice: Minimum Data Chunks With an Indivisible Header

ML domain interview

The ML round asked basic concepts, including what Naive Bayes is. The source does not preserve additional ML questions or a bounded design task.

Behavioral interview

Questions were standard, such as describing a difficulty and how it was resolved.

The candidate posted while in Team Match and did not identify the exact job title beyond PhD SWE with an ML-domain round.