The Role Map
On December 3, 2025, Shield AI and Sedaro announced a partnership to advance autonomous operations in orbit, making Shield AI's Hivemind Pilot Sedaro's preferred autonomy software for on-orbit demos — a move that pulls simulation and software engineers into a shared hiring pool across space and defense shops amid a specialist gap in autonomy talent. Christian Gutierrez, vice president of Hivemind Solutions at Shield AI, said the Sedaro tie is "a critical enabler of at-the-edge, in-orbit autonomy" (spacenews.com). The deal extends Hivemind's edge autonomy from air and sea into space.
"The combination of Hivemind's battle-proven autonomy and Sedaro's high-fidelity simulation platform will accelerate the design, simulation, and validation of autonomous behaviors for orbital missions." — Sedaro (sedaro.com)
The partnership shows up as live requisitions on Zero G Talent's board, which pulls first-party listings from company feeds. The numbers below come from that board, not third-party scrapes.
| Company | Roles added (7d) | Total open | Median salary | Sample senior role (band) |
|---|---|---|---|---|
| SpaceX | 137 | 1064 | $145k | Principal AI Engineer, Special Programs ($220k–$350k) |
| Blue Origin | 114 | 768 | $181k | Sr Principal ASIC Design Engineer ($308k–$431k) |
| Shield AI | 42 | 345 | $190k | Principal Engineer, AI/Data Platform ($320k–$490k) |
Shield AI's board lists software and autonomy posts, not hardware labor. Its careers page says engineers build cross-functional teams in AI, autonomy and robotics, shipping tech to troops in contested environments. Beyond the sample role in the table, it posted Principal Engineer, State Estimation in Dallas at $274k–$411k.
SpaceX leads the board in openings. Recent listings include Principal AI Engineer, Special Programs in Washington DC at $220k–$350k and a similar Palo Alto role needing Top Secret Clearance. A March 2026 GitHub archive shows an application for "Software Engineer, Simulations" at SpaceX, citing work on simulation infrastructure for Falcon, Dragon, Starlink, and Starship. ZipRecruiter listed 793 SpaceX simulation jobs at $44k–$94k, but the board's senior bands reflect specialist pay. SpaceX's careers page says it builds Starship for Moon and Mars and runs Starlink. A December 2026 new grad propulsion role targets 2026 or 2027 graduates at Starbase, TX.
Both giants treat simulation as core IP. Blue Origin skews silicon and software. Listings include Sr Principal ASIC Design Engineer - Terawave across San Diego, Central Texas, Kent WA, Bay Area at $308k–$431k; Sr Director NG Software Engineering in Kent at $272k–$380k; Principal ASIC Design Verification Engineer - Terawave at $269k–$376k. An Automation Engineer posting seeks a key architect for satellite constellation production automation. Blue Origin's founders aimed to enable millions living and working in space to benefit Earth.
Drone makers ride the same talent curve though the board omits their listings. The drone sector's need for simulation engineers who build virtual training environments and AI/ML engineers with drone data experience grows as operations shift from human-piloted to AI-directed, per 2026 hiring analyses. The cross-domain pull matches the Shield AI–Sedaro arc.
Hiring clusters in software, simulation, and autonomy stacks, not policy or hardware labor. The board shows active posting but not economy-wide expansion beyond these firms. That boundary sets up the case for shared toolchains.
Why Autonomy Demands Shared Simulation Stacks
The Shield AI–Sedaro partnership shows the merge across space and defense engineering, pairing battle-proven edge autonomy with orbital simulation. The same high-fidelity modeling that proves a drone can navigate without a pilot now validates a satellite’s ability to act at the edge. The computational backbone for both no longer separates.
SpaceX’s Starship landing sequence explains why. The vehicle performs a vertical touchdown via a belly-flop at near-90° angle of attack, then uses multi-engine ignition and coupled thrust-vectoring aerodynamic control to hit sub-meter precision (arxiv 2508.06520). Conventional guidance laws lean on simplified dynamics or convex optimization that linearizes the problem; those methods trade physical fidelity for speed and miss critical interactions in high-angle regimes. For a reusable launch vehicle flipping through unsteady flow, the motion history directly changes forces at a given altitude. Steady simulations and wind tunnels can’t capture separated flow near the ground.
A May 2026 test on the Flow360 platform shows the new standard. Engineers built an 80-million-node mesh and simulated 30 seconds of physical flight in 1.5 hours on 8 NVIDIA H200 GPUs, skipping the overset mesh and remeshing bottlenecks that once ate weeks. The run exposed unsteady hysteresis — data deemed essential for refining landing logic. Malhar Prajapati, on LinkedIn in May 2026, wrote that running such simulations changes how aerodynamics and controls teams at SpaceX, Blue Origin, Rocket Lab and other reusable launch firms work. The toolchain spreads before vehicles fly.
Sedaro brings high-fidelity models and a simulation environment; Hivemind Pilot brings edge autonomy from adversarial air and sea contexts. Combining them unlocks multi-agent operations where software plans and acts without ground control. Defense swarms need exactly the differentiable learned models arxiv researchers integrated into Starship trajectory loops, tightly coupled to rigid-body dynamics. The August 2025 paper said such action-conditioned transition models would underpin digital-twin virtual flight platforms, erasing the line between a rocket’s digital twin and a drone’s training sim.
Blue Origin’s January 2025 New Glenn 1 flight used AWS High Performance Computing for mission simulations and trajectory planning, with generative AI in the stack. SpaceX keeps its tool list private, but hexacoder reported in August 2025 that industry experts agree digital twin technology anchors its development, letting engineers model one engine part or the whole launch site. Digital twins cut physical test articles and let teams simulate faults pre-flight. Training workers to run these stacks bottlenecks defense projects, the same friction hexacoder flagged for aerospace.
The convergence shows in the hiring volumes on that board tallied earlier. Those postings track shared need for simulation, HPC and autonomy skills, not discrete hardware labor. The counts confirm the toolchain pull without proving at-scale training programs.
The 2026 Simulation Specialist Shortage
The drone simulator market hits about $820 million in 2026, yet a shortage of instructors who can both fly advanced drones and teach on simulators limits how fast flight schools scale, per MarkWide Research. That bottleneck sits inside a wider gap: the drone and defense sector needs simulation and autonomy engineers faster than universities and bootcamps produce them.
The pressure isn’t cyclical. Christian & Timbers called the drone sector's 2026 talent crunch structural in a March 31 analysis. The candidate pool for specialized drone roles stays thin; workers who build, run, and govern autonomous systems haven’t kept pace with demand. As drones make that transition, needs grow for such engineers and safety engineers who validate autonomous behavior, the firm noted.
The shortage traces to training lag. Regulators force the issue. The FAA Remote ID rule hit September 16, 2023; EU Regulation 2019/947 set risk-based training. Now the FAA's pending Part 108 framework pushes commercial operators from remote pilots to fully autonomous systems. MarkWide Research says FAA and EASA mandates turned simulators from optional aids into mandatory certification tools. Stricter certification needs extensive simulator hours, expanding the drone simulator market to about $2.8 billion by 2035 at a 14.6% annual growth rate. Defense agencies stay the largest end-user; private operators climb fastest.
The pipeline feeding this market cracks. Drone engineering is young; only a small subset of aerospace and robotics programs produce drone-specific graduates, Christian & Timbers reported. Most drone engineers taught themselves, switched from aerospace or robotics, or left military service for commercial roles. None scale. Worse, drone firms compete with aerospace primes, defense contractors, autonomous vehicle makers, and robotics firms for the same skills. Clearance needs scarcer defense-adjacent roles. Activity clusters in San Diego, Reno-Carson City, Phoenix, Dallas, and Northern Virginia, creating regional crunches.
Numbers quantify the squeeze. In a February 2026 survey summarized by wifitalents.com, more than half of enterprises cited skill shortages as a barrier to new tech. In 2023, almost two-thirds of organizations struggled to find workers with the right mix, and two in five said skill needs shifted within a year or two. The U.S. Bureau of Labor Statistics projects computer and math jobs to grow about 15% from 2022 to 2032, which should help, but drone-specific simulation talent stays a sliver of that pool (bls.gov).
Recruiters shift tactics. Christian & Timbers advises drone firms to hire specialist recruiters, define roles around problems not credentials, and move faster than competitors dragging candidates through six interview rounds. A March 31 piece noted top drone candidates field multiple offers at once. Pay must reflect market reality, not general surveys. Zero G Talent's board shows the premium clearly: the autonomy roles at those firms cited earlier carry six-figure bands far above broad benchmarks. Organizations using general surveys underbid and lose candidates.
Firms also upskill. Simulation works: wifitalents.com data shows time-to-proficiency for software-defined operations dropped by nearly a third after targeted simulations. Drone flight productivity rose about a fifth after scenario-based instruction. Remote pilot test pass rates averaged 86% for structured Part 107 prep versus 72% for unstructured study. Two-thirds of employees in reskilling programs reported better job performance within 6 to 12 months. Yet these fragmentary programs don’t match the broad training expansion the sector’s size implies. Regional education-industry alignment hints at responses, but evidence doesn’t confirm at-scale training pipelines.
The gap widens as the market expands. Counter-drone tech emerges as a hiring vertical; European BVLOS waivers now require simulator logs in safety cases. End-users shift from one-time simulator buys to subscription Simulator-as-a-Service, lowering costs for small operators. VR and AR platforms gain share over PC trainers.
Space and defense firms like SpaceX, Blue Origin, and Shield AI pull from the same autonomy talent pool, but the specialist gap in simulation instructors and engineers remains the brake. Board data shows hundreds of roles open weekly, yet training stays thin.
What Lies Outside the Autonomy Fence?
Zero G Talent's first-party board shows those firms holding thousands of open roles as of latest ingest. Those totals look massive next to the autonomy hires mapped earlier, but include hundreds of posts this article leaves outside the fence.
The aerospace-AI talent convergence runs through software for autonomous systems. It skips people who mill titanium, package semiconductors, or negotiate export control clauses. This section names excluded groups: hardware manufacturing labor, policy and legal hiring, and any non-autonomy role regardless of company.
Hardware manufacturing labor holds the largest out-of-scope count. Deloitte's semiconductor workforce analysis projects over a million additional skilled workers globally by 2030 to meet fab demand. HeroHunt's May 18 2026 report puts the deficit above a million and notes three in four firms couldn’t find qualified hardware engineers in 2026. US electrical engineering enrollment dropped about nine-tenths relative to computer science since the 1980s; EE degrees grew only about two-fifths from 1997 to 2020 while all other fields grew four-fifths.
Intel’s delays show the gap’s consequences. HeroHunt's May 2026 report notes Intel delayed its Ohio fab by half a decade partly because of workforce gaps; CNBC reported Feb 28 2025 the delay from 2025 to 2030 production. The company cut roughly 15% of staff, over 13,000 employees, in 2024-2025.
Aerospace production tells a similar story outside our lens. The U.S. aerospace and defense sector employs about 2.2 million workers directly and indirectly, per the 2026-05-26 TalentTraction analysis of the AIA/McKinsey 2025 study. That study found more than half of companies face skilled-trade shortages. Emprous counted over 31,000 aerospace manufacturing roles in California alone as of Jan 2026, from machinists to test engineers. A 5,000-jet delivery backlog forces OEMs to seek 123,000 more technicians over two decades, per TalentTraction analysis.
First-party board data confirms the space companies post these hardware jobs. SpaceX's recent listings include Principal Design Verification Engineer (Silicon Engineering) at $210,000–$355,000 yearly. Blue Origin advertised the same Terawave silicon role cited earlier. Those are chip roles, not simulation autonomy work. They sit on the board, but not in our tally.
We exclude policy and legal hiring for a different reason: research barely quantifies it. Federal CHIPS Act grants totaling $30.9 billion across 40 projects, per HeroHunt's May 2026 data, surely spawned compliance roles we don’t track. Security clearance delays block autonomy engineers from starting for one to two years, yet clearance adjudicators form a separate civil service pool we fence out.
Non-autonomy commercial roles round out exclusions. Shield AI's board lists Marketing Manager, Middle East at 267,000–401,000 AED and Business Development Lead at $270,000–$400,000. Those support autonomy but aren’t engineering sim posts. Drone makers’ sales and finance staff face the same cut.
The fence matters because raw headcount would drown the signal. Global AI talent demand runs three openings for every candidate, HeroHunt says, but spans all machine learning roles. Our gap is the tighter simulation and autonomy engineer subset. If you hire for factory floor or policy desk, filter Zero G Talent's full board by those categories and ignore the autonomy lens.
Open the full SpaceX or Shield AI listings and you'll see thousands of combined roles; only a sliver build the hivemind.
Working in frontier tech? Zero G Talent tracks the openings: see every open SpaceX role, browse frontier tech jobs, openings at Blue Origin and Shield AI, and the people building the field.