Impulse Space builds the vehicles that move things around once they are already in orbit: tugs, kick stages, and maneuvering spacecraft that take a payload from where a rocket drops it to where it actually needs to be. Tom Mueller founded the company in 2021. He was SpaceX’s first employee and ran propulsion there for most of two decades, and the engineering culture he is rebuilding in Redondo Beach reads like it. That matters for your prep more than any leaked question bank would, because almost nobody has written down what the Impulse interview is like yet. What we can do is reason from where the people interviewing you learned to interview.
What Impulse builds, and why it shapes the loop
Two vehicles anchor the work. Mira is an in-space maneuvering vehicle that has already flown; in 2026 two of them, LEO-2 and LEO-3, flew within about 200 meters of each other, which is a real rendezvous-and-proximity demonstration, not a press-release render. Helios is a high-energy kick stage designed to carry a payload from low Earth orbit to geostationary in under a day, with a first flight slated for 2027. There is also Electra, an electric-propulsion system that sits alongside the chemical engines. If you are interviewing here, the product is the syllabus: bipropellant propulsion, orbit determination and control beyond LEO, close-proximity maneuvering, and the flight software that ties it together.
The company is well funded for a hardware startup. Its Series D reached $808 million in September 2026, after a $500 million first close in June at a $4.26 billion valuation; the $308 million extension lifted that to roughly $5.4 billion per Reuters, citing a source close to the company. Returning backers in the extension include Lux Capital, DFJ Growth, 137 Ventures, and Valor Equity Partners, per Payload. On the government side, Impulse won a National Security Space Launch Phase 3 Lane 1 award, becoming that program’s first upper-stage prime, plus a $28 million SBIR Phase III extension tied to the VICTUS SALO responsive-space missions. That backlog is why Payload reported the new money as going mostly to hiring and manufacturing, and it is why the bar in the room is closer to SpaceX-fast than aerospace-prime-slow.
The loop, reasoning from the SpaceX lineage
Here is the inference chain, stated plainly. What is verifiable is the top of the house: Mueller was SpaceX’s first employee and ran propulsion there, and founder-led hardware startups hire early technical leads in the founder’s image, so SpaceX-style interviewing is the reasonable prior, even if I can’t confirm the rest of the staff’s backgrounds. SpaceX’s own process, from public candidate reports, runs a recruiter screen that opens with two or three technical questions, then a 30-minute technical call with a hiring manager, often a second with a senior engineer or director, and an onsite built around a short technical presentation followed by five or so one-on-one and walk-and-talk sessions with engineers and managers. A company this size, likely staffed by people who ran that playbook, would run something close to it, compressed. So expect fewer rounds than a FAANG loop and more depth per round, with at least one session that is you explaining your own past work in front of people who will push on it hard.
In practical terms, figure on roughly four to six touchpoints end to end: a recruiter screen, one or two technical calls, and an onsite, typically across two to four weeks, with the onsite usually a single day. That is the typical aerospace-startup shape, not a schedule Impulse has published, so treat it as a planning estimate.
InterviewQuery’s Impulse page, the one aggregated third-party source I found, lists behavioral, data-structures-and-algorithms, and machine-learning topics, but it reads templated (an in-space-propulsion shop is not screening most engineers on ML), so I would not weight it.
Impulse Space interview loop at a glance
| Stage (inferred from SpaceX-lineage aerospace norms and Impulse postings, Oct 2026) | What to expect | How to prep |
|---|---|---|
| Recruiter screen | Role fit and logistics, often opening with two or three quick technical questions in your discipline | Be able to state your strongest hands-on project in two sentences, with the numbers |
| Hiring-manager technical call | 30 to 45 minutes on domain fundamentals and a past project the HM pushes on | Know the physics or the system you claim to know cold; expect follow-ups, not softballs |
| Senior-engineer technical call | Deeper dive with a future teammate; propulsion, GNC, or software specifics depending on the role | Prepare one failure you debugged end to end and what you changed afterward |
| Onsite technical presentation | You present your own work (thesis, a flight hardware build, a software subsystem) to a room | Build 20 clean minutes with a real engineering decision and its trade-off at the center |
| Onsite one-on-ones | Several back-to-back sessions, some whiteboard, some walk-and-talk, mixing technical and fit | Rehearse explaining trade-offs out loud; test-fit realities and ownership both come up |
Propulsion: what the questions actually probe
For propulsion roles the giveaway is that the interviewers want to see you reason from first principles, not recite. The closest public signal comes from SpaceX propulsion loops, where reported questions include tracing how pressure, temperature, and velocity change as flow moves through a converging-diverging nozzle, and what happens to hoop stress in a pressure vessel if you thin the wall. Those are SpaceX anecdotes, not Impulse’s, but they are exactly the flavor a Mueller-built propulsion team would favor: a clean thermo or structures question you can derive at a whiteboard while narrating your assumptions. Impulse’s own senior propulsion postings ask for bipropellant experience and engine hotfire testing, so expect at least one question that starts from a test stand or a feed system and walks into failure modes rather than staying in steady-state ideal cycles.
GNC, flight dynamics, and the proximity problem
The LEO-2 and LEO-3 close approach is the most concrete GNC hook Impulse has given the public, and I would prepare as if a rendezvous-and-proximity-operations problem is on the table. Expect orbit determination and inertial navigation questions that reach beyond LEO, since Helios is a GEO-class stage, plus the control side: how you null relative velocity on approach, how sensor noise and thrust granularity set your safe standoff distance, how you reason about a trajectory when a burn slips. Impulse’s Senior Flight Dynamics and Principal GNC postings ask for seven-plus years and work across mission design, operations, and software, so these rounds lean senior and systems-level rather than toy-problem. The autonomy and perception reasoning that shows up in guides like Applied Intuition and the Waymo safety round rhymes with this: the hard part is what the vehicle does when an estimate is wrong.
Flight software and systems
Impulse’s flight-software postings prefer people who have shipped software for launch vehicles, upper stages, or in-space propulsion, and they name the stack: a real-time operating system or embedded Linux, and familiarity with NASA’s cFS, ROS2, or custom middleware. So if you are on this track, be ready to talk about determinism and timing, how you handle a watchdog and a fault response, why you would reach for cFS’s app-and-message-bus model, and how you test software you cannot physically reach after launch. A coding round is plausible; keep your C and C++ sharp, and expect embedded concerns (memory, interrupts, no dynamic allocation in flight paths) rather than a LeetCode grind. Where algorithmic questions do appear, they sit at a solid working level, so refresh the coding patterns catalog and make sure your time and space complexity reasoning is automatic, then stop; nobody here is screening for competitive-programming tricks.
Behavioral: ownership is the real screen
At a company shipping flight hardware on a backlog, the behavioral rounds are checking whether you own outcomes when they are hard, not whether you have a polished story. Prepare a project you drove through a genuine setback: a test that failed, a design you had to walk back, a schedule you had to defend or renegotiate. Concrete beats heroic. The STAR method keeps the answer from sprawling, but the content that lands is the decision you made with incomplete information and what you learned when it was wrong. Have a real why-Impulse too: in-space logistics is a specific bet, distinct from launch, and a vague “I love space” answer reads as a red flag to people who chose to build this.
Compensation, with the sourcing attached
The public reqs skew mid-to-senior: the lowest individual-contributor title I saw was Engineer II, and the Principal GNC role asks for seven-plus years, so new-grad openings are rare, though a summer propulsion internship does appear. There is no clean total-comp dataset for Impulse, so here is what is public and where it comes from. The base ranges below are California and Colorado pay-transparency disclosures pulled from Impulse’s own listings on VC portfolio job boards (DCVC, Lux Capital) between early 2025 and mid-2026, base only, each marked “plus equity”:
- Spacecraft Propulsion Engineer II (Redondo Beach): roughly $110,000 to $160,000
- Senior Propulsion Design Engineer (Redondo Beach): roughly $140,000 to $180,000
- Senior Flight Dynamics Engineer (Redondo Beach): roughly $150,000 to $200,000
- Principal GNC Engineer (Boulder, Colorado): roughly $170,000 to $200,000
- Test Automation Engineer: roughly $105,000 to $160,000
Levels.fyi, crowd-sourced and sparse, reported a median total compensation near $157,000 across roles as of April 2026, with its dedicated software-engineer page still locked for lack of submissions. So treat the median as a rough benchmark, not a number to anchor on, and treat the equity as the live variable: at a startup that just repriced from $4.26 billion to roughly $5.4 billion in a single quarter, the grant terms and strike price matter more than the base. Model your own figure with the total comp calculator, and when an offer lands, run a real salary negotiation on the equity rather than the base, because that is where a company at this valuation has room to move.
If you are shopping the broader space beat, the Vast, True Anomaly, and Blue Origin guides map the adjacent loops, and the Mach Industries guide covers the same defense-adjacent hardware-startup rhythm. Impulse’s edge in the interview is that its problem is narrow and real: moving mass that is already in space. Walk in able to talk about that problem like it is yours, and the rest of the loop is engineering you either know or you don’t.
Practice the behavioral round:
