JPL shed 1,500 people in two years — and a single Long Beach job fair proved where they're landing

The JPL Prime Contract Upheaval

NASA's decision on May 22, 2026 to compete the Jet Propulsion Laboratory management contract ended six decades of sole-source control by Caltech and opened the door to the first real institutional challenge to JPL's dominance in deep-space robotics. The current contract, worth up to $30 billion through 2028 if all options are exercised, has been Caltech's since NASA inherited the lab from the Army in 1958. Now the agency says the "rapid growth of the U.S. space economy indicates there may now be a viable competitive market" for running the place.

NASA Administrator Jared Isaacman framed the move as responsible stewardship, but the implications go beyond procurement theory. The competition will determine who runs the federally funded research and development center that operates Mars rovers, interstellar probes, and the deep-space communications backbone for NASA's planetary missions.

Caltech President Thomas F. Rosenbaum and JPL Director Dave Gallagher said in a joint memo to staff that the university "welcomes a fair and open competition" and has had a team preparing since last summer. The memo emphasized continuity for missions including FALCON, EAGLE, SkyFall, MoonFall, and GRACE-C, all targeting 2028 launches.

The Department of Energy has already run full and open competitions for five of its sixteen FFRDC management contracts over the past ten years. NASA's move follows that same precedent. The agency held an Industry Engagement Day in July 2025 to gauge interest from potential bidders. Registration and validation in SAM.gov is expected ninety days before the solicitation, which procurement analysts expect between Q4 2026 and Q2 2027.

The estimated JPL management contract value runs roughly $2.6 billion per year. Contractors expect evaluation to weigh mission continuity, facilities stewardship, workforce retention, and cybersecurity readiness. Caltech's restructuring efforts over the past year, including cost-reduction initiatives and a new contracting mechanism to scale reimbursable and philanthropic work, appear designed to position the university to compete on efficiency as well as institutional knowledge.

For the deep-space robotics workforce, the contract competition signals that JPL's monopoly on institutional knowledge is no longer guaranteed. Where that talent goes next depends on who wins and what they choose to build.

Rocket Lab's NASA Science Blitz: Sun, Earth, and Beyond

NASA picked Rocket Lab to launch two science missions (TSIS-2 and PoISIR) that previously fell under the kind of orbital-mechanics and mission-planning work that flowed through JPL. The launches will fly on Electron rockets from Mahia, New Zealand, in 2027. TSIS-2 (early 2027) carries a free-flying satellite to measure the Sun's total and spectral irradiance with two instruments covering 96% of the solar spectrum. PoSIR (no earlier than June 2027) will put two 16U CubeSats into tropical orbits spaced hours apart to track how ice content in high-altitude clouds changes daily. Both are VADR contract task orders, fixed-price indefinite-delivery awards that give NASA a 10-year pipeline worth up to $300 million across all vendors.

The missions themselves are managed elsewhere: Goddard runs TSIS-2, Vanderbilt is the PoSIR principal investigator, and Blue Canyon Technologies builds the PoSIR buses. But the launch-service award signals something bigger. NASA is routing heliophysics and Earth-science missions through a small-launcher company that hit its 70th Electron flight in August 2025 and was launching at a rate of roughly one mission every three weeks. That cadence gives science teams schedule control JPL's manifest never had to offer.

Rocket Lab's spacecraft family has already been tapped for NASA Moon and Mars missions and the first private Venus flight. When JPL alumni evaluate where deep-space mission work is actually getting done, Rocket Lab's NASA book is hard to ignore.

Loft Orbital and AI-Driven Earth Science at JPL

In June 2026, NASA's Jet Propulsion Laboratory handed over the keys to its newest AI software to a private company. Loft Orbital will host and fly JPL artificial intelligence software on its own commercial satellites, a move that outsources the lab's on-orbit processing to private infrastructure. The agreement puts JPL's Federated Autonomous MEasurement (FAME) project, funded by NASA's Earth Science Technology Office, onto Loft's AI-enabled hardware starting with flight demonstrations in June 2026, with additional deployments planned through 2028.

This is not a standard launch contract. JPL is not paying Loft to carry a sealed government instrument; it is deploying its autonomy algorithms directly onto Loft's computing architecture. FAME's core purpose is to remove humans from the Earth-observation processing loop. The software uses onboard edge computing to run deep learning classifiers and spectral analysis on acquired imagery in seconds, then autonomously retasks the satellite (what JPL calls "tip-and-queue" tasking) without waiting for ground intervention. If a satellite detects a wildfire or flood, it can task itself or cross-task other spacecraft via intersatellite links to capture follow-up images immediately.

JPL's own AI Group lists dozens of autonomy projects, including AEGIS for Mars rover targeting, MEXEC for coupled planning and execution, and CASPER for continuous replanning, all of which have traditionally run on government-built and government-operated hardware. FAME takes that same software logic and ports it to commercial platforms. The project team list makes the handoff explicit: while JPL researchers like Steve Chien and Itai Zilberstein develop the autonomy and federated scheduling algorithms, Loft engineers Jad Mogannam, Mitchell Scher, and Pieter van Duijn handle the integration and deployment on Loft's satellites. The code is JPL's; the iron is Loft's.

The commercial shift for software-defined missions

FAME's phased rollout illustrates the scale NASA envisions for commercial hosting. The project starts on 7 spacecraft by Spring 2026, scales to 20 by late 2027, and targets a stretch goal of 60 spacecraft by Summer 2028. That trajectory requires hardware standardization that government procurement rarely achieves on its own. Loft's model, buying satellites in bulk, outfitting them with high-performance processing architectures, and flying common Linux distributions as "spacecraft virtual machines," provides exactly that. JPL software gets deployed as a container with defined interfaces, tested in the cloud on virtual machines, then pushed to orbit.

Paul Lasserre, General Manager of AI for Space at Loft, said the collaboration lets JPL "process data in near-real time to support urgent decisions on Earth." That framing underplays the structural shift. When NASA's own press release notes the collaboration will "help pathfind widespread adoption of its software on commercial platforms," it is acknowledging that JPL's software-defined mission expertise is leaving the building. The engineers who write the scheduling and execution logic still sit in Pasadena, but the operators who deploy, host, and scale that logic now sit in San Francisco and Golden, Colorado.

The Reverse Brain Drain: From Pasadena to Long Beach

Over 600 JPL employees and contractors lost their jobs in the first two months of 2024 alone. 530 employees plus 40 contractors were cut in February, on top of 100 contractors let go in January. By November 2024, another 325 employees were cut. And in October 2025, 550 more received layoff notices. Across four rounds in two years, roughly a quarter of JPL's staff vanished. Director Laurie Leshin told employees the lab would stabilize at around 5,500 regular employees, down from roughly 6,500 just a few years prior.

Those numbers describe a forced migration in real time. And Long Beach noticed.

The hiring event that proved the point

On February 23, 2024, weeks after the first mass layoff, Long Beach's Economic Development department and the Pacific Gateway Workforce Innovation Network hosted a space industry job fair specifically targeting displaced JPL talent. The companies that showed up weren't small startups testing the waters. ABL Space Systems, Boeing, Relativity Space, Rocket Lab, and Vast Space all sent recruiters, alongside a dozen other aerospace firms.

Long Beach Mayor Rex Richardson said it bluntly: retaining talent in Southern California was "imperative for the success of the regional space and aerospace sector." Bo Martinez, the city's economic development director, called the moment a "rare opportunity to connect available industry talent directly with the multitude of companies looking to fill vacancies."

The firms were hiring for engineering, design, software development, project management, and operations, the exact technical mix that JPL had just shed.

DevOps and cleared engineers: the hot commodity

The talent flowing out of JPL isn't generic. Two skill profiles carry outsized demand in commercial space right now: DevOps engineers and TS/SCI-cleared technical staff.

Zero G Talent's board shows Rocket Lab added 43 roles in the past seven days. Several map directly to these categories: DevOps Engineer II/Senior roles in both Long Beach and Littleton, Colorado, paying $115,000–$170,000 and $125,000–$170,000 respectively, plus Senior Systems Engineer and Senior Flight Software Engineer positions that explicitly require TS clearance, with salary bands reaching $190,000.

That's not coincidental. JPL's software-defined mission expertise, the infrastructure that runs autonomous deep-space operations, is built by exactly the kind of engineers who configure Kubernetes clusters, manage Terraform deployments, and hold active clearances for classified NASA and Department of Defense collaborations. When those engineers hit the market, commercial firms don't just gain a competent DevOps hire. They gain someone who has already operated software in environments where a bug costs a billion-dollar mission.

The Long Beach–Pasadena corridor is roughly 30 miles. A JPL engineer can change employers without changing school districts. That geographic proximity turns what might be a slow talent bleed into a fast siphon.

The pattern is unambiguous. JPL's institutional knowledge in mission-operations software, orbital-mechanics simulation, and classified payload integration is walking out the gate at 4800 Oak Grove Drive and driving south on the 710 to Long Beach, or increasingly to Rocket Lab's Colorado office and the other commercial sites now competing for the same cleared, DevOps-fluent engineers who once had no reason to leave Pasadena.

Blue Origin's Lunar Permanence and the Deep-Space GN&C Talent Grab

Blue Origin's Lunar Permanence business unit, the division building the Blue Moon lunar lander family, has become the most visible commercial sink for the kind of deep-space Guidance, Navigation, and Control talent that JPL once hoarded. The unit's job postings read like a JPL organizational chart flipped inside-out: GN&C Flight Controls Engineer, Senior GN&C Systems Engineer, Sr GN&C Hardware Engineer, Navigation Hardware Engineer, Motor Control Avionics Engineer, Systems Engineer for Fault Management. All of them sit at 8082 Space Commerce Way, Merritt Island, FL, the same Blue Origin campus that serves as the company's lunar-lander development hub, a short drive from Kennedy Space Center.

The work is not Earth-orbit satellite bus engineering. The Lunar Permanence GN&C roles explicitly cover rendezvous, proximity operations, docking, entry, descent, and landing, the high-stakes control problems JPL refined over decades on Mars Science Laboratory, Juno, and the upcoming Mars Sample Return campaign. The GN&C Flight Controls Engineer posting asks for hands-on experience with control allocation, reaction control systems, thrust vector control, and 6-DOF integrated simulation, skills that map directly onto JPL's flight-controls heritage. The Sr GN&C Hardware Engineer role demands deep knowledge of IMUs, star trackers, lidar, radar, and GPS for space navigation, plus experience with thermal vacuum, shock, and radiation testing of flight hardware. These are not entry-level positions. The minimum qualifications call for 8+ years of relevant experience and U.S. citizenship or permanent residency.

The salary ranges tell their own story. GN&C Flight Controls Engineers in Merritt Island pull between $150,931 and $211,303 annually; the same role in the Greater Seattle Area tops out at $230,512. The Sr GN&C Hardware Engineer role lists $156,802 to $219,522 for Washington applicants. Blue Origin also offers 401(k) matching up to 5%, stock options, and up to four weeks of paid time off, a benefits package that competes with or exceeds what a GS-scale government research position offers, especially when factoring in equity upside.

Zero G Talent's own board data underscores the pace: Blue Origin added 153 roles in the past week alone, with GN&C Flight Controls, Avionics Harness Design, and Instrumentation & Controls positions all feeding the Lunar Permanence pipeline. That's a hiring velocity that matches or exceeds JPL's annual output of new GN&C hires.

The geographic clustering matters. Blue Origin's Space Coast campus sits inside a corridor that already includes SpaceX's Cape Canaveral operations, Northrop Grumman's Melbourne facilities, and a growing cluster of smaller firms like Aegis Aerospace and Stoke Space, all of which list GN&C and flight-controls roles on the same job boards. An engineer with a TS/SCI clearance and five years of JPL-adjacent GN&C experience can now move from Pasadena to Merritt Island, or from JPL's Mars mission division to a Blue Moon lander team, without changing the fundamental nature of the work. That mobility is the threat to JPL's institutional depth, not any single contract loss, but the steady optionality that commercial deep-space teams now offer to the same talent pool.

What This Means for the Deep-Space Workforce

The deep space robotics market will grow from $1.51 billion in 2025 to $2.21 billion by 2030. That expansion tracks a direct transfer of institutional knowledge from government labs to private companies. When Intuitive Machines paid $30 million to acquire KinetX in October 2025, it bought deep-space navigation expertise that once lived exclusively inside NASA's contractor base. Similar talent pull-through is happening across the sector right now.

NASA, by contrast, posted 11 roles (mostly technician positions) during the same period Rocket Lab added 43 and Blue Origin added 153.

Where the demand concentrates

The talent migration is not uniform. It targets specific skill sets that legacy institutions no longer monopolize.

• GN&C and autonomous navigation: As missions push past low Earth orbit, guidance, navigation, and control engineers who can operate with communication delays of minutes or hours command the highest premiums. Blue Origin's Lunar Permanence salaries reflect this scarcity.
• Cleared DevOps and infrastructure: Rocket Lab's Long Beach postings for TS-cleared DevOps and flight software engineers show that commercial firms now handle classified mission work once reserved for traditional defense primes.
• AI-based mission software: The deep space robotics market's fastest-growing software segment covers autonomy, path planning, and sensor fusion, precisely the capabilities Loft Orbital is now contracting from JPL's own AI teams.

The leading indicator

Stanford HAI's 2026 AI Index Report documents a 20% drop in employment for software developers ages 22 to 25, even as AI-related productivity in software development hits 26%. The pattern in deep-space robotics runs in the opposite direction, toward experienced, cleared engineers who can adapt legacy mission-planning expertise to commercial hardware cycles. This is not a field where junior generalists compete. The premium falls on people who have already guided hardware through the radiation, thermal, and latency constraints of cislunar space.

The companies winning these engineers are the ones with funded missions on the manifest. Firefly Aerospace's Blue Ghost Mission 1 proved a private company could land on the Moon and operate autonomously for 14 days. Intuitive Machines bought KinetX to secure the navigation stack for its next lunar attempts. Rocket Lab's NASA science launches require orbital-mechanics talent that understands both the physics and the clearance requirements.

JPL will keep building rovers. But the commercial deep-space workforce is being assembled on the Space Coast and along the 710 corridor, job by job and clearance by clearance, at salary bands the civil service scale was never designed to match.

Working in space? Zero G Talent tracks the openings: browse space jobs, openings at Blue Origin, Rocket Lab and NASA, and the people building the field.