Anduril Software Engineer (New Grad) Interview Experience
Anduril · Entry level · Software Engineer
The weirdest round was a HackerRank about arranging two football teams for a photo, and the interviewer basically admitted it was confusing on purpose. I left thinking I had definitely failed it, then still got pushed to finals.
Interview process
I went straight to a recruiter screen with no OA, then did a 60-minute engineer phone screen, and after that the final was three separate 60-minute rounds. The biggest theme was ambiguity: the phone screen HackerRank was intentionally confusing, the other technicals kept expanding the problem, and I had to ask a lot of clarifying questions instead of expecting a neat prompt. The behavioral with a senior leader was also tougher than usual because he went deep on every line of my resume and kept circling back to why Anduril and why defense. The turnaround was fast and I usually heard back the next day, and I honestly thought I had failed the phone screen but still got moved forward. I ended up not getting the offer, but the process made it pretty clear they care about fundamentals, clean code, and whether you actually align with what the company does.
Interview rounds · 5
- 1
Recruiter screen
BehavioralThis was a straight recruiter screen with no OA first, and it felt less like a generic script and more like him trying to figure out what I actually wanted to work on and whether I really meant it with Anduril.
Q1. What do you want to work on here: perception, backend infra, ML infra, or ML?
How they answeredI walked through my past projects and the kinds of problems I actually wanted to spend time on, because he seemed to be mapping me to possible areas already. It felt like he wanted a real signal on whether I leaned more toward things like backend infra, ML infra, or something else, not just a vague 'I'm open to anything' answer.
Follow-up questions- Based on your projects and past experience, what kinds of work are you most interested in?
Q2. Why Anduril, especially when it's not a popular company for everyone?
How they answeredI said I was interested in the impact of the problems and that it felt like a high-growth company with a lot of hard things to work on. Under the surface, it definitely felt like he was checking whether I was actually okay with defense and whether I had a real reason for choosing Anduril over the other places on my resume.
Follow-up questions- Why Anduril instead of the other companies you've worked at before?
- 2
Technical round
CodingData Structures & AlgorithmsThis was a 60-minute HackerRank coding screen, and honestly it was one of the weirdest rounds I've had because the prompt was confusing on purpose and the engineer seemed to care more about whether I kept clarifying constraints than whether I gave some perfect polished solution.
Q1. You're trying to take a photograph of two teams arranged in rows. How would you order them so shorter players are in front and taller players are behind?
How they answeredI spent most of the round clarifying what the prompt even meant. My read was that there were two separate team arrays, I could sort within a team but not mix teams, then decide which team goes in front so nobody in front blocks someone shorter behind. After that he added uneven team sizes and empty spaces, so I had to rethink the arrangement and consider impossible cases too. He barely pushed on complexity. It was more like, keep asking questions and get it to work.
Follow-up questions- Which team should be in front and which should be in back?
- What if the teams are different sizes and there are empty spaces?
- What if it's impossible to arrange them so no shorter player gets blocked?
- 3
Final / onsite round
BehavioralThe behavioral final was with a very senior leader, and it felt more like getting grilled on my actual track record and motivations than doing canned STAR answers.
Q1. Walk me through your resume.
How they answeredHe basically went line by line through my resume and pushed a lot harder than most behavioral interviewers do. I answered which experience I liked most, whether I got return offers, how I thought I performed, and what my manager would probably say about me. It didn't feel like standard behavioral prep really mapped to it. It felt like he was trying to figure out if I could back up every line on my resume and what I was actually like to work with.
Follow-up questions- Which past role did you like the most?
- Did you get a return offer?
- How would you rate your performance there?
- What would your manager say about you?
Q2. Why do you want to work at Anduril?
How they answeredThis one felt like an actual conversation, not just a checkbox. I talked about the scale and impact of the problems and why the company's growth interested me, but he kept pulling on it in a way that felt like he was testing whether I genuinely wanted to work in defense. As a new grad, that was honestly hard, because they're really probing for values alignment, not just a polished company answer.
Follow-up questions- What about your previous experience makes Anduril the right next step?
Q3. Which of my teams or areas sounds most aligned with your background?
How they answeredI talked through the kinds of teams that matched what I'd done before and what I wanted to work on next. It almost felt like early team matching. After that, he spent a lot of time telling me about his org, how the company had grown, and why he'd stayed so long, which made the round feel very fit-oriented and pretty personal.
- 4
Final / onsite round
CodingData Structures & AlgorithmsOne of the final technicals started like a normal coding problem, but they kept expanding the setup, so the real test was whether I kept asking the next sensible question instead of locking into the first version.
Q1. Find the shortest distance between two points.
How they answeredI started with the basic case in 2D space and talked through the simple Euclidean distance setup first. Then the interviewer added obstacles, so now I had to think about routes that couldn't just go straight through. The base problem wasn't that hard. The point was that they wanted me to refine the solution step by step as the assumptions changed. I also talked through time and space complexity because that's just my habit.
Follow-up questions- What if the points are in space?
- What if there are obstacles you can't pass through?
- 5
Final / onsite round
TechnicalCodingThe other final technical was more object-oriented design than algorithm grinding, and the interviewer really cared about code quality and refactoring, not just whether I could hack out a working answer.
Q1. Design a doubly linked list.
How they answeredI had to design the class and write methods like insert, delete, and search. The interviewer cared a lot about whether the code was clean, functional, and organized. I first had a bigger insert flow, then refactored it into separate cases like insert at the beginning, end, and middle because he kept pushing on cleanliness and edge cases. In hindsight, it was a fundamental question, but I wasn't expecting something that basic and I got tripped up more than I should have.
Follow-up questions- Implement methods like insert, delete, and search.
- How would you handle insert at the beginning, end, and middle?
- Can you refactor this to make it cleaner?
Tips from the candidate
I'd still do a lot of LeetCode, especially tagged questions, but I wouldn't only prep the usual array and graph stuff. I'd also review really basic design questions like linked lists, queues, and deques, because I got caught off guard by how fundamental one of the finals was. Know your resume inside out, because the behavioral can get into return offers, manager feedback, and what you actually did, not just canned 'tell me about a time' stories. And go in expecting weird open-ended prompts, so ask clarifying questions early and don't assume the question is clean on purpose.
Company culture
I got the sense they hire pretty fast, and even as a bigger company they were moving with next-day turnaround a lot of the time. For new grad, the loop felt fundamentals-first because team matching happens later, so they weren't testing some super specific domain skill upfront. The engineers seemed to like ambiguity and wanted me to ask questions instead of just jumping into code, and a lot of them seemed very C++-heavy even though I did my interview in Python. They also kept coming back to why Anduril and why this industry without saying 'defense' super directly, so mission and values fit are definitely part of the bar, not just technical skill.