NEOM's Tonomus has activated its Oxagon data centre, planted 150 operational staff in an 18,000-square-metre facility on the Red Sea coast, and signed a $5 billion agreement with DataVolt to build a 1.5-gigawatt AI campus — all while The Line construction remains suspended. The smart city is paused; the AI infrastructure machine underneath it is not. The H1 2026 hiring picture at NEOM is therefore a split-screen: construction headcount collapsed 74.6 percent between January and December 2025, from 342 active postings to 87 (per ENTRA job posting tracker and LinkedIn Talent Insights data), while a distinct and smaller category of senior AI, cognitive systems, and data infrastructure roles is expanding into territory that no Gulf employer has mapped before.
This is not the same as Riyadh's Humain intake or Aramco Digital's industrial-AI cohort in Dhahran. NEOM is hiring for a live operating environment — an AI platform, branded NEOS, that must simultaneously run 200 million IoT devices by 2030, manage autonomous mobility in a city that was designed with zero legacy infrastructure, and maintain data sovereignty within Saudi jurisdiction. The role categories this creates are not on any standard job ladder in San Francisco or London. They are being defined in Tabuk Province, and the engineers who fill them will have built the category from scratch.
The Roles NEOM Is Hiring For
The NEOS cognitive platform is Tonomus's central operating system: it aggregates data from the Oxagon district's energy systems, logistics networks, IoT sensors, and security infrastructure into a predictive layer that is supposed to anticipate operational needs rather than respond to them. Running this platform at scale requires a class of engineer that Tonomus CEO Joseph Bradley has called a "cognitive city architect" — a hybrid between a senior ML platform engineer and a distributed-systems infrastructure specialist, operating in an environment with no legacy codebase and no comparable prior deployment.
The Oxagon DataVolt campus — 300-megawatt Phase 1 targeting 2028 operational readiness, scaling to 1.5 GW total — adds a second tier of specialist demand: MLOps engineers and AI infrastructure reliability engineers who can operate GPU cluster workloads at a scale that has no regional precedent. NEOM's Humain joint venture with xAI added 500 megawatts of committed capacity on top of the DataVolt deal, and Humain is separately targeting 6.6 GW of its own data centre capacity by 2034. The engineering headcount required to operate those facilities is not being filled by fresh graduates. Tonomus and Humain are competing for senior infrastructure engineers with verifiable hyperscale experience — profiles that in 2026 are employed at AWS, Google Cloud, or Microsoft Azure, and are being recruited directly.
Three role clusters define the active H1 hiring demand at NEOM:
Cognitive platform engineers — senior ML engineers with distributed inference and real-time data pipeline credentials, responsible for the NEOS operating layer. Tonomus's public technical documentation sets the target state at 60 locations, 1,800 structures, and 200 million IoT devices. The engineers building toward that state are not operating on a product roadmap with quarterly milestones. They are constructing the category specification and the platform simultaneously. There is no comparable civilian deployment to reference.
AI-urban systems architects — engineers with cross-domain expertise spanning autonomous vehicle path-planning, computer vision, and edge-compute systems. Oxagon is designed for autonomous mobility from the first paving stone, which means there is no fallback to conventional traffic management infrastructure. The path-planning and object-recognition systems must work at a site that is actively under construction and expanding. The engineering problem is real-time and evolving.
Data sovereignty specialists — a role category specific to Saudi Arabia's AI regulatory context and one that has no direct equivalent in the UAE. Saudi Arabia's Global AI Hub Law, enacted and effective 1 August 2026, establishes data localisation requirements mandating that sensitive government and national infrastructure data remain within Saudi jurisdiction. Tonomus, as the data and AI operator for NEOM's physical environment, requires engineers who can architect compliant data pipelines, certify AI workloads against the new regulatory framework, and manage the boundary conditions between NEOM's private cloud and external compute providers. This is a compliance-engineering speciality with almost no trained talent supply and no comparable private-sector precedent in the region.
What NEOM Pays for These Roles
The compensation architecture for senior NEOM hires in 2026 sits above the Riyadh standard-market band. For experienced engineers at the ML platform and AI infrastructure level — three to seven years of relevant experience, ideally from hyperscale environments — Tonomus packages in H1 2026 are tracking SAR 45,000 to SAR 65,000 per month in base salary, with performance bonuses in the range of SAR 80,000 to SAR 120,000 annually (per ENTRA Salary Survey Q2 2026 and recruiter conversations). At the top end, principal-level cognitive infrastructure roles are reaching SAR 85,000 monthly base.
Saudi Arabia applies zero personal income tax on wages and salaries. A Tonomus senior engineer at SAR 55,000 monthly base with a SAR 100,000 performance bonus retains approximately $202,667 in total first-year cash — before the relocation stack is added. The NEOM relocation structure for qualifying international hires includes a SAR 30,000 to SAR 80,000 one-time relocation allowance, full housing coverage at Tabuk Province equivalent (SAR 84,000 to SAR 96,000 per year), annual return flights, full medical for employee and family, and a school allowance for dependent children that NEOM's published recruitment materials (reviewed by ENTRA, April 2026) cite at up to $60,000 per child per year. The KSA Premium Residency — specifically the Exceptional Competence Residency product, which NEOM-affiliated employers cover as part of the senior offer standard package at SAR 100,000 ($26,660) — provides sponsor-free residency with no kafala dependency, covering the employee and family unit under a single renewable permit.
The Premium Residency mechanism is structurally significant for senior international candidates. Unlike the UAE Golden Visa, which requires meeting category thresholds and periodic requalification, the Exceptional Competence Residency is issued on a per-individual professional-merit basis and is portable across employers within the Kingdom. For a mid-career engineer currently on an H-1B in the United States facing renewal uncertainty, or an ex-DeepMind hire in London on a Skilled Worker visa tied to a single employer, the sponsor-free character of the Exceptional Competence Residency removes a structural constraint that otherwise limits negotiating leverage at the point of counter-offer.
Competing With Abu Dhabi for the Same People
NEOM is recruiting from a talent pool that G42 and Core42 are also working hard to lock. The competition is direct and documented. G42's ML engineer band on Levels.fyi is reporting AED 360,000 to AED 408,000 annually for mid-level roles; data scientists at G42 are landing AED 360,000 to AED 473,000. Stargate UAE — the $30 billion AI campus in Abu Dhabi being developed by G42, OpenAI, and Oracle, with 200 megawatts of the first phase going live in 2026 — is pushing senior AI engineer compensation above AED 80,000 per month for infrastructure leads, per recruiter conversations tracked by ENTRA.
Both Saudi Arabia and the UAE apply zero personal income tax. The UAE adds a 0% income corridor across both ADGM and DIFC structures and has established the AI Specialist Visa as a dedicated residence instrument for AI professionals. NEOM's counter-position runs on three axes.
First: mission specificity. G42 and Core42 are building compute infrastructure in Abu Dhabi — important work, but functionally similar to what AWS or Azure engineers do at scale. NEOM's cognitive platform engineers are building something that does not exist anywhere else on earth: an AI operating system for a city that was designed with no analogue constraints. The greenfield argument is genuine, not marketing. For a senior engineer motivated by category-creation rather than production-scale optimization, NEOM is offering a problem that G42's Reem Island data hall cannot replicate.
Second: data sovereignty mandate. The Global AI Hub Law creates a regulatory floor under Tonomus's data sovereignty roles that the UAE does not currently replicate. Engineers working on NEOM's data localisation stack are not doing discretionary compliance work — they are implementing mandatory national infrastructure with direct regulatory standing. That role has structural job security, government backing, and national-priority classification that insulates it from the budget-cycle volatility that affects private-sector AI infrastructure spending.
Third: the Oxagon compute base. The DataVolt $5 billion campus, combined with the Humain-xAI 500 MW commitment and Humain's 6.6 GW 2034 target, makes Oxagon one of the highest-density planned AI compute sites globally. Engineers who join the operational team for Phase 1 (targeting 2028) are positioning for careers at the centre of Saudi Arabia's sovereign compute infrastructure for the next decade — not as a contractor rotation but as permanent, residency-anchored staff in a site where the capital commitment makes the work durable.
The gap in NEOM's pitch — and recruiters working the senior IC corridor between London and Riyadh know this — is site reality. Oxagon is an industrial port under construction in a remote stretch of Tabuk Province. The Tonomus Telecommunications Center is operational; the DataVolt campus does not exist yet beyond its first foundation work. Candidates who accept NEOM offers in H1 2026 are accepting a construction-site reality before the product reality arrives. G42's engineering teams in Abu Dhabi are working in a fully built, globally connected city with a UAE Golden Visa in hand and a two-hour flight to any European capital. For senior candidates with families and options, NEOM's school allowance and housing provision — while competitive — are compensating for a location premium that not every candidate prices at zero.
What H2 2026 Looks Like
Tonomus's internal hiring roadmap, per sources with knowledge of the planning timeline, projects the senior AI engineering headcount at the Oxagon facility to roughly double between H1 and H2 2026 as the DataVolt Phase 1 construction progresses and NEOS platform deployments expand across the operational Oxagon district. The roles going active in Q3 and Q4 are weighted toward the compute infrastructure layer — MLOps platform engineers, GPU cluster reliability specialists, and network architects for the AI campus — rather than the cognitive platform build that dominated H1 openings.
The Humain-NEOM interface is the structural factor most likely to reshape the hiring picture. Humain, the PIF-owned sovereign AI company operating separately from Tonomus, has committed to co-locating infrastructure at Oxagon and is running parallel hiring into AI infrastructure leadership. Whether Humain absorbs the senior engineering pipeline that would otherwise go to Tonomus, or whether the two entities develop complementary but non-competing talent pools, will determine whether NEOM-as-employer functions as one brand or two in the H2 recruiting market. Based on ENTRA tracking of LinkedIn activity and executive search mandates in Q2 2026, they are currently competing for the same profiles from the same search firms.
The Global AI Hub Law's procurement-certification requirement comes into full operational scope in Q3 2026 (effective 1 August 2026). Data sovereignty specialists hired in H1 will be the first cohort to qualify for the regulatory certification tracks that the Law creates. For the small number of engineers in the global talent market who combine AI systems architecture credentials with data-localisation and sovereign-cloud compliance experience, NEOM is not one employer among several. It is the only employer running a live deployment at this regulatory specification, at this infrastructure scale, with this level of sovereign-capital backing — and the window to enter as a foundational team member rather than a later-stage hire closes as the DataVolt campus moves from permitting to operation.
Global Career PlatformFind AI talent. Find your next role.
Booking is hotels. · Airbnb is apartments. · ENTRA is global careers.