The interview process at a robotics company is not the same as the interview process at a general software company. Whiteboard LeetCode is rare. System design is common but different (real-time constraints, hardware in the loop, sensor fusion tradeoffs). Take-home assignments are basically dead. Onsite loops are becoming shorter but more targeted.
Below is what candidates report for the top robotics employers, plus what job postings themselves tell us. Take the specifics as a starting point, not gospel. Every team's process varies.
Anduril
Rough shape: recruiter screen, hiring manager conversation, one or two technical interviews, then an onsite (Costa Mesa, Playa Vista, Austin, or Denver).
What Anduril actually filters for: engineering excellence, comfort with defense-adjacent work, ability to move fast in a mission-driven organization.
Timeline: 3-5 weeks. Fast when they want you.
What the postings signal: 12% of Anduril robotics postings mention onsite, 7% mention system design. Actual reports suggest onsite happens for most senior roles.
Common myths:
- "Anduril requires US citizenship." Only 0-2% of postings explicitly require it. Green card holders eligible for most roles.
- "It's a defense contractor culture." Not the traditional prime-contractor culture at all. Growth-stage tech company that happens to serve government.
- "You need weapons systems experience." False for the vast majority of roles. Autonomy, perception, and simulation roles hire heavily from AV, robotics research, and general ML backgrounds.
Preparation tips:
- Know what Lattice is. Anduril assumes candidates understand the product.
- Read up on the current specific programs they hire for (Bolt, Fury, Roadrunner, etc.). Interviewers appreciate context.
- Direct, clear communication. Anduril's culture punishes unfluffy answers less than most companies, but rewards brevity.
Zoox
Rough shape: recruiter screen, hiring manager, technical phone screens, and an onsite in Foster City. Multiple technical interviews often specifically testing hardware/software integration.
What Zoox filters for: engineering depth in a specific area, ability to work on a full-stack robotics problem (perception + planning + controls + hardware), and comfort with Amazon-owned but engineering-led culture.
Timeline: 4-6 weeks.
What the postings signal: 15% of Zoox postings mention system design specifically. This is one of the highest rates in the catalog for robotics employers. Onsite mention is low (2%) but candidate reports suggest most senior roles do get onsite loops.
Common myths:
- "Amazon owns Zoox now so it's the Amazon interview." False. Zoox retained their engineering culture and their interview process is much more technical and less script-driven than typical Amazon.
- "You need SAE Level 5 autonomy experience." No. Perception, planning, controls, or hardware experience from any autonomous vehicle context is welcome.
Preparation tips:
- Zoox does deep system design. Be prepared to reason about sensor placement, planning latency budgets, and safety-critical software patterns.
- Have specific opinions about tradeoffs in AV systems. Vague "it depends" answers don't land.
- Know their vehicle. Zoox is the last remaining "purpose-built robotaxi" story in the US.
Boston Dynamics
Rough shape: recruiter, technical phone screens, and an onsite in Waltham for most roles.
What Boston Dynamics filters for: strong engineering fundamentals, physical intuition, and cultural fit with a legacy engineering organization now under Hyundai ownership.
Timeline: 4-8 weeks. Boston Dynamics is not fast.
What the postings signal: system design mentioned in 10% of postings. Onsite mentioned rarely but candidate reports say it's standard.
Common myths:
- "They only hire PhDs." Absolutely false. They hire many strong engineers without PhDs.
- "Boston Dynamics doesn't ship." Half true. Spot does ship in real quantities. Atlas is still R&D-heavy. Stretch is between the two.
- "Hyundai has made it corporate and boring." Interview reports suggest the engineering culture is very much still Boston Dynamics.
Preparation tips:
- Physical intuition matters. Be able to reason about dynamics, control loops, and physical constraints in real time during interviews.
- Have specific opinions about their products. Interviewers can spot people who watched the YouTube videos but don't know the engineering underneath.
- Fundamentals over trendy skills. If you know controls theory deeply, that outweighs knowing the latest ML framework.
NVIDIA (robotics roles)
Rough shape: recruiter, hiring manager, multiple technical interviews, often ending in a "hiring committee" review that adds meaningful timeline.
What NVIDIA filters for: technical depth in a specific area, ability to work with GPU-accelerated systems, and pedigree signals (top schools, top prior employers). More than most robotics companies, NVIDIA weights resume prestige.
Timeline: 4-8 weeks. Hiring committee adds delay.
What the postings signal: system design mentioned in 4% of postings. That's low relative to the other big players. Onsite mention very low.
Common myths:
- "NVIDIA is a chip company; they don't hire roboticists." Wrong. Isaac Sim, DRIVE, and Cosmos alone employ hundreds of robotics engineers.
- "You need CUDA expertise." Useful but not required for many roles.
Preparation tips:
- Know what Isaac Sim and Cosmos are. NVIDIA assumes candidates understand their robotics product landscape.
- Have opinions about GPU-accelerated robotics. Their culture rewards genuine curiosity about hardware acceleration.
- Prestige signals matter more here than at other robotics companies. If you have a top program on your resume, lead with it. If you don't, lead with specific accomplishments.
Skydio
Rough shape: recruiter, technical phone screens, and onsite in San Mateo for most roles.
What Skydio filters for: technical depth in drone-relevant areas (perception, autonomy, controls, hardware), and comfort with a smaller-team culture where individual output matters a lot.
Timeline: 3-5 weeks.
What the postings signal: 12% of Skydio postings mention onsite specifically. System design at 7%.
Common myths:
- "Skydio is a drone company." Consumer drone was 5+ years ago. Now they're heavily defense and enterprise.
- "You need drone experience." No. Perception, controls, autonomy from any context transfers well.
Preparation tips:
- Understand the pivot from consumer to defense/enterprise. Don't come in expecting to work on consumer camera drones.
- Small-team dynamics. Skydio is much smaller than Anduril; each engineer has more surface area.
Wayve
Rough shape: recruiter, technical phone screens (usually 2-3), and onsite in Sunnyvale (or London for UK-based candidates).
What Wayve filters for: exceptional ML research background, systems-thinking about autonomy, and cultural fit with a research-forward but production-focused organization.
Timeline: 4-6 weeks.
What the postings signal: 10% of Wayve postings mention onsite. System design at 7%.
Common myths:
- "Wayve is a UK company so US roles are second-tier." Absolutely false. Their US engineering team is treated as core, not adjunct.
- "You need to know end-to-end learned driving specifically." No. Any strong ML research background with an interest in driving is welcome.
Preparation tips:
- Read the Wayve papers before your interview. Their approach is distinctive and they expect candidates to have engaged with it.
- Be able to reason about safety and edge cases. Wayve interviewers push hard on how you handle rare events in learned systems.
Waabi
Rough shape: recruiter, technical phone screens, and mostly-virtual loop with occasional Toronto onsite.
What Waabi filters for: research-heavy ML backgrounds, foundation-model interest, and comfort with a smaller team.
Timeline: 4-6 weeks.
What the postings signal: system design at 7%. Onsite mentions rare.
Common myths:
- "Just a Canadian version of Wayve." Waabi's approach is genuinely different: simulator-first and much more focused on autonomous trucking.
Preparation tips:
- Know the difference between simulator-first and reality-first approaches to AV.
- Understand Urtasun's research history. Her prior work at Uber ATG is the foundation of Waabi's approach.
General observations
Three things across interview reports for the last twelve months.
System design interviews are becoming standard at robotics companies. Zoox at 15%, Anduril at 7%, Boston Dynamics at 10%, Wayve at 7%. If you're a senior engineer, expect to run through system design at any of these companies. Robotics system design is different from web-scale system design; be prepared to reason about real-time constraints, sensor placement, and safety.
Take-homes are basically dead in robotics. Only 0-1% of postings mention take-home assignments. This is a hardware and integrated-system industry where take-homes don't test the things employers care about.
Coding challenges are rare. LeetCode-style interviews happen but far less than in general software. Robotics companies mostly want to see you reason about robotics-specific problems, not solve algorithm puzzles.
What actually gets people offers
Two patterns repeat across candidate reports.
Specific relevant project experience. If you've built something adjacent to what the company builds, put it at the top of your resume. Interviewers at these companies care about specific relevant experience much more than general prestige signals.
Being able to reason through a problem out loud. Robotics interviews reward "show your work" thinking. If your default style is silent-until-the-answer, work on this. The industry values reasoning over answers.
Methodology
I combined three sources: 1) text search across active postings for specific interview-format language, 2) publicly reported interview experiences on Glassdoor, Blind, and Reddit, 3) notes from candidates I've talked to over the last six months. Specific processes vary by team, hiring manager, and role level.
If you've interviewed recently and your experience differs from what's described here, email hello@greatrobots.ai. Corrections welcome.