Google PhD Early Career AI/ML Loop: Rightward Grid Paths, Research, and Team Match

Role: Software Engineer, PhD Early Career — AI/ML
Result: Entered Team Match
Report time: March 2025
Application: Applied online

Timeline

• October 2024: applied online
• Early December: recruiter outreach
• January 8, 2025: phone screen
• January 15: advanced to virtual onsite
• January 31 and February 3: four onsite rounds
• A little over one week later: Team Match discussions began

Phone screen

The coding problem was a variation of Top K Frequent Elements. The variation and follow-ups were not preserved, so the standard problem is not asserted as an exact match.

Virtual onsite coding 1

The candidate described a problem solved with topological sorting plus graph BFS, but the source's linked prompt did not render and the actual task is missing.

Virtual onsite coding 2: count rightward grid paths

Start at the bottom-left of a board and reach the bottom-right. Every move must go right, right-up, or right-down. Return the total number of valid paths.

The interviewer asked for both DFS and dynamic-programming approaches.

Practice: Count Rightward Grid Paths

ML and behavioral rounds

The ML interviewer discussed the candidate's research papers and asked quick questions connected to YouTube work. The exact questions were not remembered and did not reach full system-design depth.

The behavioral round covered ordinary teamwork and conflict stories. One coding round was not completed, but recruiting still began Team Match while feedback review was ongoing.