Interview experiences

DoorDash L4 Software Engineer Onsite Interview (NYC)

DoorDashL4 Software Engineer·New York, NY (interview conducted virtually)·Interviewed October 2022Outcome unknown

Updated July 17, 2026

A candidate interviewing for a mid-level L4 Software Engineer position at DoorDash in New York shared a detailed account of a fully virtual process, running from an initial phone screen through a four-round onsite loop. The loop combined a data-structure design problem, a classic algorithm question, and two conversations covering both hiring-manager topics and a system design problem.

The write-up describes a wide range in interviewer engagement across the loop: one round asked for a specific implementation approach and would not accept alternatives, another repeated the same behavioral questions under a different label, and the final system design round left most requirements-gathering questions unanswered. No final outcome was reported, but the notes give a clear picture of what each round actually covered.

How the process went

  1. Phone Screen

    A single coding problem was given over the phone screen: counting how many location clusters are fully covered, given a matrix of driver-available locations and a matrix of driver-needed locations.

  2. Onsite Loop

    A fully virtual onsite with four back-to-back rounds: a data-structure design problem, a hiring-manager conversation, an algorithm question, and a combined system design and domain-knowledge round.

  3. Outcome

    The write-up ends after describing the final round and does not state whether an offer followed.

Phone Screen

Coding / matrix problem

  • Given a matrix of driver-available locations and a matrix of driver-needed locations, determine how many location clusters are fully covered by drivers who can service them.

Onsite Round 1: Design File System

Data structure design (trie)

  • Implement an in-memory key-value store where values are stored at slash-separated paths, supporting get(path), set(path, value), create(path, value), and delete(path) — similar to the 'Design File System' problem with added set and delete operations.

The interviewer stated a trie-based solution was required and would not consider other approaches.

Onsite Round 2: Hiring Manager Chat

Behavioral

  • Describe a project at your current company.
  • What technical challenges came up, what did you learn, and what would you do differently?
  • Describe a time you had a conflict with someone over how to approach something.
  • Describe a time you had to argue for one side of a tradeoff.

Onsite Round 3: Algorithm

Coding

  • Largest Rectangle in Histogram (the standard problem, unmodified).

Onsite Round 4: System Design and Domain Knowledge

System design

  • A repeat of the hiring-manager-style questions about a current project, technical challenges, and what would be done differently, framed under a 'domain knowledge' heading.
  • Design a donation application for a three-day charity event expected to process roughly $100 million across ten charities, integrating with a third-party payment processor that accepts card details through a simple REST endpoint and returns a success response, with all collected funds held in a single account for manual disbursement after the event.

The candidate described the interviewer as largely unresponsive to requirements questions and visibly distracted for most of the discussion.

Key takeaways

  • For data-structure design questions phrased like 'Design File System' with added write operations, practicing a trie-based implementation ahead of time can help, since some interviewers will insist on that specific approach.
  • Classic algorithm problems such as Largest Rectangle in Histogram can show up unmodified in onsite loops, so fluency with well-known patterns still pays off.
  • Having a small set of reusable behavioral stories ready (a project overview, a technical challenge, a conflict, and a tradeoff) is useful, since the same questions can resurface across more than one round.
  • In system design rounds where the interviewer declines to clarify requirements, moving forward with clearly stated assumptions and continuing to narrate tradeoffs out loud keeps the conversation productive even without feedback.
  • Interviewer engagement can vary a lot round to round within the same loop, so it is worth treating one disengaged interviewer as a single data point rather than a signal about the whole process.

Practice a DoorDash interview

Rehearse out loud against the kinds of questions in this story — with an AI interviewer that asks follow-ups.

Practice this interview

Source

The questions and process facts come from the candidate's public write-up, linked below. The retelling above is our own summary.

Candidate's public write-up on LeetCode Discuss