Waymo data scientist candidates report a recruiter screen, a technical assessment or screen, and a multi-interview loop covering statistics, experimentation, Python coding, ML and behavioral questions.
$187K
Avg. Base Comp
$278K
Avg. Total Comp
Not reported
Typical Rounds
Not reported
Process Length
Candidates describe the Waymo data scientist interview as tied closely to the team's real work, with statistics and experimentation appearing most often. One candidate said much of the loop covered statistical tests, including the assumptions behind each and when to apply it. Another described a first round on experimentation and a second on sampling to compare two algorithms. A third spent a technical round modeling a problem out loud.
Coding is usually Python, but the style varies by report. One candidate saw a dynamic programming problem and a numpy question on sampling 3D bounding boxes. Another was asked to sample data from a unit disk and work with dataframes. A third described a statistics-focused Python problem as easy to medium.
The loop can also include ML and data-system topics. Reports mention K-means at a conceptual level, how transformers work, long-tail imbalance, a dataset-based case study and a data system design interview. Candidates also describe questions about past projects and about why you want to work on self-driving or at Waymo. None of these candidates received an offer, so prepare for breadth and be ready to explain your own project work clearly.
Synthesized from 8 candidate reports by our editorial team.
Had an interview recently?
Share your experience. Unlock the full guide.
Real interview reports from people who went through the Waymo process.
4 rounds: the first interview was data intuition covering experimentation, the second was also data intuition covering sampling to compare two algorithms, the third was a coding interview in Python covering dataframes, and the fourth was a behavioral interview focusing on experience. Every interview asked things about my experience, though.
Questions asked: The coding question was to sample data from a unit disk. I can't remember a ton of the data intuition questions.
Share your own interview experience to unlock all reports, or subscribe for full access.
Sourced from candidate reports and verified by our team.
Topics based on recent interview experiences.
Featured question at Waymo
Write a function to return the value of the nearest node that is a parent to both nodes.
| Question | |
|---|---|
| Car Recommendation Architecture | |
| Location Frequency | |
| Pathfinder in Maze | |
| Minimum Days for Scheduling All Meetings | |
| Shortest Path Algorithms | |
| Azure Kubernetes Infrastructure | |
| Statistically Significant Test | |
| Empty Neighborhoods | |
| 2nd Highest Salary | |
| Employee Salaries | |
| Top Three Salaries | |
| Last Transaction | |
| First to Six | |
| Top 3 Users | |
| Experiment Validity | |
| User Experience Percentage | |
| Distance Traveled | |
| Merge Sorted Lists | |
| First Touch Attribution | |
| Fair Coin | |
| Third Purchase | |
| Button AB Test | |
| Download Facts | |
| Get Top N Frequent Words | |
| Normal Distribution Sample | |
| String Shift | |
| Size of Joins | |
| 500 Cards | |
| Find Bigrams |
Synthesized from candidate reports. Individual experiences may vary.
Candidates report one or two recruiter calls covering their background, an overview of the role and why they are interested in self-driving or Waymo. One candidate also described walking through a prior project. Prepare a concise motivation answer and project summary.
One candidate reports a technical assessment before a four-round onsite. Another had a data-fluency conversation that felt more like a chat, where they described a past project, their role, technical choices and outcome. Prepare to explain your project decisions clearly.
Candidates report interviews centered on statistics: which statistical tests to use and their assumptions, experimentation, sampling to compare two algorithms, and modeling a problem aloud. Some also discussed their background in depth. Expect to justify methods rather than recite definitions.
Reported coding questions include dynamic programming, sampling from a unit disk, dataframe work and numpy-style sampling of 3D bounding boxes. One candidate called the Python problem easy to medium. Practice both algorithmic problems and practical implementation in Python.
Some candidates report ML fundamentals such as K-means, transformers and long-tail imbalance, a dataset-based case study, data system design, and behavioral or hiring-manager discussion. These topics were not reported by everyone, so treat them as possible rather than guaranteed.
