← Back to Directory

Nuro interview notes, page 1 of 4

Mid-sized

Most of these Nuro write-ups come from Software Engineers (23), then Machine Learning Engineers (9) and Backend Engineers (2). The most reported rounds are the technical phone screen (17), the system design round (13), the behavioral / leadership round (2) and the hiring manager screen (1).

35 interview notes · updated Sep 2026

Nuro·Machine Learning Engineer·Technical Phone Screen

Apr 2026
Coding screen for an MLE role at Nuro. One question, friendly vibe, but the interviewer pushed back on my first approach and made me redo it from scratch.
  • Implement an LRU cache. The interviewer initially accepted a high-level solution but then asked you to implement it using a doubly linked list instead of a built-in ordered map.

“My first instinct was to use OrderedDict and call it done.”

View Post

Nuro·Software Engineer·Technical Phone Screen

Apr 2026
Nuro coding round, one algorithmic problem about clustering points by proximity. The follow-up on optimizing for large sparse inputs was where things got interesting.
  • Given N points in 2D space and a distance threshold k, group all points such that any two points within distance k of each other (transitively) belong to the same cluster. How would you approach this, and how would you optimize it when N is very large but the points are spread out sparsely?

“The union-find part I got pretty quickly, just iterate over all pairs, check distance, union if below k.”

View Post

Nuro·Machine Learning Engineer·Technical Phone Screen

Apr 2026
Second round coding interview at Nuro for an ML engineer role. Pretty standard algorithmic problem but the implementation detail tripped me up a bit.
  • Given n children each with some value, find the maximum. The twist is you need to track the top two maximums using a single loop.

“My first instinct was to just sort and grab the top two, which they pushed back on immediately.”

View Post

Nuro·Software Engineer·Technical Phone Screen

Apr 2026
Nuro SWE interview threw a graph problem dressed up in autonomous vehicle flavor. Basically LC 269 with a coat of paint, which I only realized mid-solve when the adjacency structure clicked.
  • Given a list of priority-ordered observations of autonomous delivery vehicles, derive a single consistent global ordering of vehicle priorities across all observations, or report that no valid ordering exists.

“It's LC 269 wearing a disguise.”

View Post

Nuro·Machine Learning Engineer·Technical Phone Screen

Apr 2026
Interviewed for an MLE role at Nuro and got a classic weighted random sampling problem. Pretty standard stuff if you've seen it before, but the implementation details can trip you up if you're not careful.
  • Given an array of weights, implement a function that randomly picks an index where the probability of selecting each index is proportional to its weight.

“I knew the prefix sum approach going in, build the cumulative array in the constructor, then on each pick generate a random number and binary search for the right bucket.”

View Post

Nuro·Software Engineer·Technical Phone Screen

Apr 2026
Nuro SWE interview with a meaty concurrency design problem. Not your typical LeetCode session, this one was all about threading models and real-world scheduling tradeoffs. Came out of it feeling like I'd either crushed it or completely missed the point.
  • Design a Job Scheduler class that runs a user-provided function on a background schedule at a specified frequency in hertz. The class should support scheduling and descheduling, and you need to handle concurrency, drift correction, overrun behavior, and clean shutdown.

“This took up basically the whole interview.”

View Post

Nuro·Machine Learning Engineer·Technical Phone Screen

May 2026
Interviewed for an MLE role at Nuro and got a fairly classic sparse matrix problem. Nothing too wild but the complexity analysis at the end tripped me up more than I expected.
  • Given two sparse matrices A (m x k) and B (k x n) as 2D arrays, compute the matrix product A * B. Optimize for sparsity by using a compact representation and only processing non-zero entries. Then analyze the time complexity in terms of nnz(A) and nnz(B).

“I went straight to building a row-to-list-of-(col, val) map for A and a similar structure for B.”

View Post

Nuro·Backend Engineer·Technical Phone Screen

May 2026
Nuro backend interview with two meaty coding problems back to back. The grid reveal was manageable but the scheduler design went deep fast, covering concurrency, cancellation, and overload handling in ways I wasn't fully prepared for.
  • Given a hidden Minesweeper board and a click position, implement the reveal logic using BFS.
  • Design a thread-safe in-memory periodic job scheduler with register, cancel, start, and stop operations.
  • How would you handle long-running tasks, cancellation, clock drift, and backpressure in the scheduler?
  • How would you prevent the same job from running concurrently with itself?
  • How would you test the scheduler deterministically?
  • How would you persist scheduled jobs across process restarts?

“Felt okay about this one.”

View Post

Nuro·Software Engineer·Hiring Manager Screen

May 2026
Interviewed for an EM role at Nuro. Just one question I can remember clearly, and it's the kind that sounds easy until you're actually in it.
  • How do you decide when to delegate something versus keeping ownership yourself?

“I fumbled this a bit.”

View Post

Nuro·Backend Engineer·Onsite - System Design / Architecture

May 2026
Two back-to-back system design rounds at Nuro for a backend role, both pretty heavy on distributed systems fundamentals. One was a webhook delivery platform, the other a vehicle telemetry aggregation system. Left feeling like I did okay but probably undersold some of the tradeoffs.
  • How would you design a multi-tenant webhook delivery platform?
  • How would you design a telemetry ingestion and aggregation system for a fleet of connected vehicles sending measurements every few seconds?
  • Why is exactly-once webhook delivery basically impossible in practice?
  • If one customer's endpoint is consistently failing or slow, how do you prevent it from affecting the rest of the system?
  • How do you handle out-of-order telemetry events in a vehicle metrics pipeline?
  • What storage would you use for serving recent dashboard data versus storing raw historical telemetry long-term?
  • If you change your aggregation logic, how do you replay historical telemetry to backfill the new results?

“This one took me a minute to scope properly.”

View Post