Skip to main content
defense

Anduril Is the Only Company Building Both the Drone and the Network That Commands It — and 138 Job Postings in Seven Days Prove It

By Andrew Chang

A Foundation, Not a Finish Line

The U.S. Army handed Anduril the keys to its most important modernization program on June 22, 2026, selecting the company to lead the common data layer baseline for Next Generation Command and Control. The announcement marks the first major production contract in NGC2's shift from prototyping to continuous delivery across Army formations, and it positions Anduril's Lattice platform at the center of a system the Army has compared to an AI-powered brain for the entire service.

NGC2 is the Army's answer to Joint All-Domain Command and Control, the Pentagon's push to fuse data across land, air, sea, space, and cyber into a single decision-making architecture. Right now, the Army runs its command-and-control on 17 separate programs of record that don't share data cleanly. NGC2 replaces that patchwork with a four-layer technology stack: transport, infrastructure, data, and applications. The data layer is the linchpin. Without it, nothing else connects.

Anduril will provide the tactical edge of that data mesh through Lattice, working with Palantir's Foundry as the cloud backbone and Raft's Data Platform for registries, transformation, and federation. The work falls under Anduril's existing 10-year, $20 billion enterprise licensing agreement with the Army, so no separate dollar figure was disclosed. Joseph Welch, the Army's portfolio acquisition executive for command and control, said the service is "already moving out with the converged data layer architecture."

The selection validates a prototyping run that started in 2025, when the Army awarded Anduril a $99.6 million Other Transaction Authority agreement to build an NGC2 prototype for the 4th Infantry Division. Over near-monthly Ivy Sting exercises at Fort Carson, the team scaled the system from roughly 65 Tactical Edge Computers to 130, and from 10 soldier-connected devices to over 2,500 at the Ivy Mass combined exercise in May. Lockheed Martin ran a parallel prototyping track with the 25th Infantry Division in Hawaii using a different architecture called C2 Fix. Both tracks will now converge under the Anduril-led common data layer, though Lockheed retains its lead on 25th ID's full-stack operational implementation.

Brig. Gen. Shane Taylor, the capability program executive for Command and Control Information Network, called the award "a major step forward as NGC2 evolves into a phase of continuous delivery." Dan Driscoll, Secretary of the Army, told the Senate Armed Services Committee in May that the service needs capabilities in all 11 divisions within two to three years, not the five-year window the current plan models. The bottleneck, he said, is not technical. It is "simply a spending pacing item."

What that means in practice: the Army is betting that a common data layer built on commercial infrastructure, rather than a custom government-developed system, can compress the path from experimentation to fielded capability. The workforce required to deliver on that bet is where things get interesting.

Costa Mesa: The Unlikely Epicenter of Military-AI Integration

Anduril's headquarters in Costa Mesa, California, a beach-adjacent Orange County city better known for shopping malls than defense contracting, has quietly become the densest concentration of military-AI integration engineering in the country. The reason sits in the job postings.

Open roles on Anduril's own careers page and across its LinkedIn listings show a company hiring aggressively in its home base for positions tied directly to the Connected Warfare team, the division modernizing command and control for U.S. and allied forces. Zero G Talent's board data shows 138 Anduril roles added in the past seven days alone, with Costa Mesa listed as the location for multiple Connected Warfare positions, including a Senior Technical Program Manager role paying six figures.

The titles tell the story. "Mechanical Engineer, Connected Warfare" and "Mechanical Engineer, Production, Connected Warfare" (both based in Costa Mesa) are not generic defense postings. The production role asks for someone who will "own the mechanical definition" of products being built at scale and shipped to customers worldwide. The standard Connected Warfare mechanical role wants four or more years of professional experience, hands-on fabrication skills, and the ability to design hardware that survives battlefield conditions. Both list salaries in the $129,000 to $171,000 range, before equity.

These aren't prototype-shop jobs. The production posting explicitly calls out managing a bill of materials in a PLM tool, implementing design-for-manufacture feedback, and overseeing engineering documentation for final as-shipped products. That language (ownership of a product in medium- to high-volume production, root cause analysis, permanent corrective action) is the vocabulary of a company moving from demo units to serial manufacturing. It's the NGC2 common data layer translating into hardware that has to work in the field, at scale, not just in a lab.

Anduril also lists a Generative AI Integration Engineer in Costa Mesa, a role focused on how AI agents interact with classical software systems inside Lattice OS, the operating system that powers the company's family of defense platforms. The posting describes work on agent tool calling, prompt engineering, and test infrastructure for agentic interaction patterns. It's a software-side complement to the hardware roles, and it points to the same underlying shift: the NGC2 win isn't just a contract line. It's pulling in people who can bridge the gap between AI models and the military systems those models are supposed to command.

The Costa Mesa cluster matters because of its proximity to a broader Southern California defense-tech corridor. Competitors like Mach Industries (Huntington Beach), True Anomaly (Long Beach), and Raytheon (Fullerton) are within a short drive, and their own LinkedIn job listings show overlapping mechanical and manufacturing engineering roles. But Anduril's concentration of Connected Warfare roles, spanning mechanical design, production engineering, manufacturing engineering, and AI integration software, is distinct in its focus on the integration layer itself, not just the platforms sitting on top of it.

This is what the NGC2 common data layer looks like in hiring terms: a mid-sized city in Orange County filling up with engineers whose job is to make autonomous warfare systems talk to each other, survive field conditions, and ship in volume.

Why Anduril Now Sits Alone at the Intersection

Anduril is now the only defense-tech company simultaneously building production hardware for the Air Force's drone wingman program and architecting the Army's central data backbone. That overlap is creating a workforce demand profile that doesn't exist anywhere else in the sector.

The Air Force's June 17 decision to award Anduril a full production contract for the FQ-44A Fury alongside General Atomics' FQ-42A put the company in an unusual position: it went from prototype to manufacturing in under two years, faster than any new fighter program since the 1970s. Mark Shushnar, Anduril's vice president for autonomous airpower, said the company had already implemented full-rate production processes and tooling on prototype aircraft before the contract was signed, identifying issues during prototyping specifically to smooth the manufacturing transition.

That production line needs to deliver at least 150 combat-capable CCAs by the end of the decade, with the Air Force's long-term target sitting at roughly 1,000. The service requested $996.5 million in procurement funding for fiscal 2027 alone, plus $150 million in advance funding, the first time it has asked Congress to buy CCAs rather than just develop them. Unit costs are targeting under $30 million per aircraft, roughly one-third the price of a Lot 17 F-35A.

At the same time, Anduril is leading the Army's NGC2 common data layer effort, which means the same company is now responsible for both building the semi-autonomous platforms and wiring them into the network that connects sensors, shooters, and commanders across domains. The workforce implications compound fast. Engineers who understand Lattice are needed for both programs. So are systems integration specialists, manufacturing engineers for airframe production, and program managers who can navigate two different acquisition timelines with two different services.

No other defense-tech company faces this exact double exposure. General Atomics is building the FQ-42A but isn't leading a parallel Army data layer program. Lockheed and Northrop lost the CCA competition and are competing on autonomy software, not airframes. Anduril is the only one that has to staff for manufacturing scale on the Air Force side while simultaneously building the integration architecture the Army will use to connect its entire next-generation command-and-control stack. That makes its hiring pattern a leading indicator for where military-AI integration employment is actually heading, not toward more prototypes, but toward production lines and the people who run them.

What the Job Postings Actually Reveal

Anduril's open roles are a blueprint for what the NGC2 common data layer actually demands, and the picture that emerges is less "software company with a defense vertical" and more "hardware-software integration shop that has to ship physical systems at wartime speed."

The clearest signal is EagleEye, Anduril's wide-area surveillance system. The company is hiring both a Senior Manufacturing Engineer and a Lead Manufacturing Engineer for the product, both based in Costa Mesa, both posted in the past several weeks.

Role Base Salary Range Location
Mechanical Engineer, Connected Warfare $129,000 – $171,000 Costa Mesa
Mechanical Engineer, Production, Connected Warfare $129,000 – $171,000 Costa Mesa
Senior Manufacturing Engineer (EagleEye) $146,000 – $194,000 Costa Mesa
Lead Manufacturing Engineer (EagleEye) $146,000 – $194,000 Costa Mesa

These are not prototyping positions. The job descriptions focus on New Product Introduction, the transition from prototype to high-volume production, along with tool design, process creation, vendor sourcing, and quality planning. Anduril needs someone who can take a complex electromechanical assembly and build the factory line that produces it repeatedly, at scale, under the documentation standards defense contracts require.

The preferred qualifications tell the real story. Both EagleEye listings call for hands-on experience with R/C vehicles, autonomous flight vehicles, or flight controllers. Anduril is not looking for a generic manufacturing engineer who happens to have worked at a medical device company. It wants someone who has personally built or overseen the assembly of autonomous hardware, someone who has soldered, machined, or fabricated physical systems with their own hands. The listings explicitly say personal or extracurricular project work counts. That is a deliberate filter for people who understand what happens when a design hits the gap between CAD and a working unit.

The Lead role adds a management dimension: define workflows across cross-functional teams, build and mentor a manufacturing engineering team. Anduril is standing up a manufacturing organization, not filling a seat.

What Anduril is not hiring for is just as telling. There are no pure marketing roles, no policy fellows, no business development generalists in this batch. The NGC2 win has forced the company into a production posture. The common data layer is software, but it runs on hardware that has been manufactured, tested, deployed, and maintained in the field. Every open manufacturing and process engineering role is a direct consequence of that reality.

The Israel Hub and the Export Question

Anduril is in late-stage talks with Israeli officials to establish a local defense-tech operation, a move that would turn the NGC2 common data layer from an American program into an international one. The company has shortlisted candidates to run Israel operations and is in the final stages of appointing one of them, according to reports from Globes and Calcalist. The role would cover sales to Israel's Ministry of Defense and R&D integration for Anduril's global product line.

The groundwork was laid in February 2026, when founder Palmer Luckey made a quiet two-day visit to Israel, meeting Prime Minister Benjamin Netanyahu and senior defense officials. The trip, organized by Lux Capital founder Josh Wolfe, included meetings with 10 Israeli defense startups arranged through the Directorate of Defense Research and Development (DDR&D), among them Smart Shooter, Kela, Oz, Skana Robotics, Regulus, Magnus Metal, eyesAtop, and AriEV. Calcalist reported that Anduril has already signed agreements with Israeli companies, including ASIO, which is expected to supply components for its unmanned aerial systems, and has expressed interest in LiteVision, a drone camera startup backed by 8VC.

Yitzhak Applbaum, chairman of the Kinetica defense-tech fund, said Anduril is evaluating entry through both organic growth and acquisitions. "Israel is the most logical place for Anduril to be," Applbaum told Calcalist, citing the company's interest in investing in Israeli startups, acquiring companies, and eventually selling to Israel's defense industry and the IDF.

The timing is not accidental. Anduril closed a $5 billion funding round at a $61 billion valuation in mid-2026, roughly double its valuation from the prior June. International expansion supports the company's stated goal of an IPO at a significantly higher valuation. Anduril's 2025 revenue hit $2.2 billion, up from roughly $1 billion the prior year, and its headcount has grown from 90 in 2019 to approximately 7,000. An Israel hub would give the company access to one of the world's most dense concentrations of combat-tested defense engineering talent, engineers who have fielded AI-assisted targeting, autonomous interceptors, and electronic warfare systems under real operational conditions.

But the export path runs through two of the world's most restrictive regulatory regimes simultaneously. Anduril must navigate International Traffic in Arms Regulations (ITAR) on the American side and Israel's own strict export control laws on the other. Any NGC2-derived technology shared with or developed in Israel would need to clear both, a process that can take months or kill a deal outright. Anduril's London office already serves as its European engineering and business development hub, and the company launched a joint venture with UAE-based EDGE Group in November 2025 to manufacture VTOL drones. Israel would be a different proposition: a sovereign customer, not just a manufacturing partner, with its own defense industrial base and competing local startups that already hold government contracts for AI-assisted surveillance and loitering munitions.

The workforce implications are concrete. If Anduril establishes an Israeli operation tied to NGC2 development, it would need systems engineers who understand both the Lattice platform and Israeli operational requirements, integration engineers who can bridge U.S. and Israeli data standards, and program managers fluent in both countries' export control frameworks. That's a narrow talent pool. Anduril's current international expansion has been limited to the London office and the Irish edge computing acquisition, neither of which involved the regulatory complexity of a dual-use technology hub in a country that is simultaneously a close U.S. ally and an independent defense exporter.

The Israel talks signal that Anduril's leadership views the NGC2 common data layer as a product, not just a contract. A program built for the Army that can also serve allied defense establishments is a program that scales beyond a single customer. Whether that vision survives contact with ITAR will determine if the Israel hub becomes an engineering center or stays a sales office.

Can Lockheed and RTX Keep Up?

The NGC2 win doesn't just give Anduril a contract. It gives the company a recruiting weapon. Anduril's board on Zero G Talent lists 138 roles added in the past seven days, a pace that dwarfs what legacy primes are advertising for comparable integration work. The question is whether Lockheed Martin and RTX can compete for the same engineers when their own growth is pulling them in different directions.

Lockheed's current hiring energy is spread across a wide surface area. The company is quadrupling production of the Precision Strike Missile, expanding HIMARS output, and pushing into hypersonics and the Golden Dome missile defense program. On the software side, Lockheed is leading a European industry bid for NATO's Next Generation Modelling and Simulation competition, built on its ACES platform, a 30-year-old simulation ecosystem that Raashi Quattlebaum, vice president of land and maritime solutions, said would "strengthen interoperability and readiness across the Alliance." That's a real capability, but it's also a different problem than what NGC2 demands. ACES connects simulation environments for training and planning. NGC2's common data layer has to fuse live sensor feeds, autonomous platforms, and command decisions in real time on a battlefield. The engineering overlap is partial at best.

RTX's Raytheon unit is in the middle of a production surge that leaves little room for the kind of software-integration hiring Anduril is doing. Five framework agreements with the Department of War aim to push Tomahawk cruise missile output from roughly 60 per year toward 1,000 annually, alongside scaled production of AMRAAM, SM-3, and SM-6 interceptors. A $384 million contract modification for Tomahawk extends work to January 2029. That's a manufacturing and supply-chain hiring story (factory technicians, quality engineers, procurement specialists), not a bid for the machine-learning and real-time-systems engineers that NGC2's data layer requires.

The structural advantage is this: Lockheed and RTX are hiring to deliver platforms. Anduril is hiring to deliver the connective tissue between platforms. Engineers who want to work on the system-of-systems problem, the thing the Army has failed to build for two decades, now have a clear destination. Anduril's 138 fresh listings suggest the company knows the window is open and the talent pool is finite.

Watch the next six months. If Anduril's Costa Mesa headcount keeps climbing while Lockheed's simulation bid and RTX's missile lines consume their recruiting bandwidth, the NGC2 win will have done more than build a data layer. It will have redrawn the map of who builds the next generation of American warfare, and who just supplies the parts.

From Common Data Layer to Autonomous Kill Web

The NGC2 common data layer baseline the Army just established isn't the end state. It's the foundation for something the service has been building toward for years: a connected kill web where sensors, shooters, and autonomous systems share targeting data across every domain without human bottlenecks.

The architecture already points there. NGC2's four-layer stack (infrastructure, transport, data, and application) was designed so that 90% of the Army operates at the application layer, pulling from a unified data layer that synchronizes information across warfighting systems in real time. COL Paul Smith, writing in Army Sustainment, described it as a "clean sheet" approach built for large-scale combat operations where speed of decision matters more than hierarchy. Once data flows through a common layer, the next step is letting algorithms act on it.

Project Convergence Capstone 6 in July 2026 is the immediate milestone. The Army will run a division-scale force-on-force NGC2 validation at the National Training Center, testing whether the common data layer holds under contested conditions. Brig. Gen. Shane Taylor, the capability program executive for C2, called it the event that lets the Army "rapidly advance from prototyping to scaling product delivery." If PC-C6 validates the architecture, the path opens to fielding NGC2 components across multiple divisions, not just the 4th ID and 25th ID test units.

That scaling is where the workforce implications compound. Anduril's role as lead for the common data baseline means its Lattice platform becomes the connective tissue. Palantir's Foundry handles edge-to-cloud data mesh. Raft provides data registries and federation tools. Each additional division that comes online multiplies the integration surface, with more sensor feeds, more autonomous platforms, and more edge compute nodes that need to talk to each other through the same data layer.

The sustainment dimension adds another layer of complexity. NGC2's data-informed sustainment decision-making capability, which uses AI and machine learning to forecast resupply needs, optimize route planning, and adapt to adversary drone activity, depends entirely on the common data layer functioning as designed. Logistics teams get predictive resupply planning only if the data layer ingests real-time consumption rates, casualty estimates, and battle damage assessments from the same feed that drives fires and maneuver decisions. That integration work requires engineers who understand both the data architecture and the operational problem.

For hiring signals to watch in 2026, three areas stand out. First, look for Anduril and its NGC2 partners to expand forward-deployed engineering roles, the Costa Mesa-based positions that put engineers alongside operational units during exercises like Lightning Surge and Ivy Mass. These roles feed soldier feedback directly back into the development cycle, and the Army's public affairs office said those two divisions will "quickly begin to implement NGC2 common components" through their operational implementation leads. Second, watch for roles tied to Project Convergence Capstone 6 specifically (integration engineers, test leads, and data platform specialists who can validate the system at National Training Center scale). Third, track sustainment-adjacent positions as the Army's Sustainment Warfighting System gets woven into the NGC2 application layer. That's where predictive logistics meets the common data layer, and it's a workforce category that barely existed two years ago.

The common data layer is the prerequisite. Everything the Army wants to do with autonomous systems, collaborative combat aircraft, robotic logistics, and AI-driven fires, runs through it. The companies and engineers building that layer now are building the backbone of how the U.S. military fights for the next two decades.


Working in frontier tech? Zero G Talent tracks the openings: browse frontier tech jobs, openings at Anduril Industries, and the people building the field.

Ready to Start Your Space Career?

Browse defense jobs and find your next opportunity.

View defense Jobs