The US technology sector is experiencing a profound structural shift that extends far beyond the headline layoffs of recent years. Two powerful forces are converging to reshape employment prospects for Indian H-1B visa holders: massive capital expenditure on artificial intelligence infrastructure—particularly data centres, GPU procurement, and model training—is redirecting corporate spending away from hiring additional engineers, while simultaneously, Trump-era policy changes are raising the cost and complexity of H-1B sponsorship itself. For the Indian diaspora working in or seeking to enter the US tech industry, understanding this dual squeeze is essential to navigating career decisions over the next 18–24 months.
The implications are neither uniformly catastrophic nor uniformly positive. Rather, they create a bifurcated labour market: certain technical specialisms—particularly those centred on AI infrastructure, security, and applied research—remain in acute demand and command premium compensation. Meanwhile, roles that were historically reliable pathways for mid-level Indian engineers—full-stack development, quality assurance, operations—face structural headcount pressure. For NRIs already in the US, the challenge is strategic upskilling. For those abroad considering immigration, the calculation now includes higher visa costs and stricter wage-floor requirements that may price out earlier-career candidates from smaller sponsors.
The AI Capex Reallocation: Why Tech Companies Are Spending on Chips, Not Headcount
Over the past two years, the largest US technology companies have redirected unprecedented sums toward artificial intelligence infrastructure. This is not discretionary marketing spend or R&D budget reallocation; it is foundational capex—the construction of data centres, procurement of graphics processing units (GPUs), and investment in the computational backbone required to train, fine-tune, and deploy large language models and agentic AI systems at scale.
The scale is staggering. Industry analysts estimate that major cloud and software companies are collectively spending tens of billions of dollars annually on AI infrastructure. For NRIs evaluating their career trajectory, the critical insight is this: that capital is not flowing to traditional software engineering headcount. Instead, it is flowing to semiconductor supply chains, data-centre construction, and a smaller, more specialised cohort of engineers focused on infrastructure, systems optimisation, and model deployment. A company that might have hired 500 mid-level software engineers two years ago to scale a monolithic application may now hire 50 infrastructure engineers and 20 ML operations specialists—and deploy AI agents to automate tasks that would previously have required 200 additional engineers.
This shift has immediate consequences for visa sponsorship. Employers facing pressure to justify H-1B hiring under stricter scrutiny now face an awkward question: why hire a mid-level full-stack engineer on H-1B when the company's growth strategy no longer requires scaling headcount in that category? The answer, increasingly, is that they do not—or they do so only for roles where no domestic candidate pool exists. For NRIs, this means that roles once considered stable and repeatable are now subject to structural, not cyclical, headcount pressure.
Trump-Era H-1B Policy Changes: Higher Costs, Stricter Wage Floors, Enforcement Tightening
Concurrent with the AI capex shift, US immigration policy has moved toward measures that increase the cost and complexity of H-1B sponsorship. While the H-1B programme itself remains in place, recent policy changes—including fee increases, wage-based selection mechanisms, and stricter enforcement of prevailing-wage rules—have altered the economics of hiring foreign workers.
Fee structures have risen. Employers now face higher filing fees and, in some cases, additional fees tied to company size and workforce composition. For smaller and mid-sized tech firms—precisely those that have historically been reliable sponsors for Indian engineers—these costs represent a meaningful percentage of the salary budget allocated to foreign hiring. A company that previously could sponsor five H-1B workers at a given cost may now sponsor three at the same total outlay.
Wage-based selection mechanisms have also tightened. Under prevailing-wage guidance, employers must demonstrate that they are paying H-1B workers at or above the prevailing wage for the occupation and geographic area. This is not new in principle, but enforcement has become more rigorous, and the definition of "prevailing wage" has shifted upward in many jurisdictions. For NRIs, this creates a paradoxical situation: the policy is ostensibly designed to protect US workers by preventing wage undercutting, but it also means that employers are less likely to sponsor H-1B candidates for junior or mid-level roles where the prevailing wage may exceed what the company would pay a domestic entry-level hire. Senior and specialised roles, by contrast, command higher prevailing wages that employers are already paying to domestic candidates, making sponsorship less economically disruptive.
For NRIs seeking to enter the US market, or those on earlier visas seeking renewal, this environment is materially more challenging than it was five years ago. The combination of higher visa costs and stricter wage enforcement means that employers must make a stronger business case for each H-1B hire. That case is easiest to make for roles where labour-market data shows genuine scarcity—which, in the current environment, means AI infrastructure, security, and senior applied-research positions.
Which Roles Are Most Exposed: The Structural Decline of Mid-Career Generalist Positions
Not all tech roles face equal pressure. Understanding which categories are most exposed—and which remain resilient—is essential for NRIs planning their career moves.
Roles Under Structural Pressure
Mid-level full-stack development has been a traditional pathway for Indian engineers entering the US market. These roles—building and maintaining web applications, microservices, and customer-facing features—are precisely the kind that companies are now automating or consolidating. AI code-generation tools, combined with the shift toward agentic AI that can handle routine development tasks, have reduced the marginal value of hiring additional mid-level full-stack engineers. For NRIs in these roles, the risk is not immediate redundancy, but rather stagnation: fewer new positions opening, slower promotion velocity, and reduced leverage in salary negotiations.
Quality assurance and testing roles face similar pressures. Test automation has long been a trend, but AI-powered test generation and anomaly detection are accelerating it. Companies are consolidating QA teams and shifting remaining QA engineers toward more strategic, exploratory testing roles that require deeper domain knowledge. For NRIs in traditional QA roles, this means that the pathway to career progression is narrowing unless they upskill into areas like test infrastructure, security testing, or performance engineering.
Operations and DevOps roles are experiencing a more nuanced shift. Cloud-native infrastructure and Kubernetes have already automated much of the work that traditional ops teams performed. Now, AI-driven observability, automated remediation, and self-healing systems are further reducing headcount requirements. NRIs in ops roles need to evolve toward platform engineering, infrastructure-as-code specialisation, or security operations to remain competitive.
Roles Remaining in Acute Demand
AI and machine-learning infrastructure engineers remain in severe shortage. These roles involve designing and optimising the systems that train and serve large language models, managing GPU clusters, optimising data pipelines, and building the tooling that enables other teams to work with AI effectively. Employers face genuine difficulty hiring for these roles domestically, which makes the H-1B business case straightforward. For NRIs with expertise in distributed systems, systems programming, or data infrastructure, this is a growth area.
Security engineers and architects are also in persistent demand. As AI systems proliferate, the attack surface expands, and companies are investing heavily in security infrastructure, threat detection, and compliance. This is not an area where AI has yet meaningfully reduced headcount; if anything, it has increased it. For NRIs with security expertise, visa sponsorship remains relatively accessible.
Applied research and senior individual contributors in AI and machine learning remain highly sought after. Companies are willing to sponsor H-1B visas—and to explore alternative visa categories—for researchers and senior engineers who can contribute to model development, novel architectures, or breakthrough applications. The prevailing-wage requirement is less of a barrier here because these roles command high salaries regardless of visa status.
Upskilling Pathways for NRIs: Concrete Steps Toward Resilience
For NRIs currently in the US tech sector, or those planning to enter it, the strategic response is targeted upskilling. The goal is to move from roles facing structural headcount pressure toward specialisms where demand remains acute and visa sponsorship remains economically rational for employers.
AI Infrastructure and Systems Engineering
This is the highest-priority upskilling area. NRIs with software engineering backgrounds can transition into AI infrastructure by deepening their knowledge of distributed systems, GPU programming, and machine-learning frameworks. Concrete steps include: contributing to open-source projects in the AI infrastructure space (PyTorch, TensorFlow, Ray); pursuing certifications or specialised courses in GPU computing and distributed systems; and seeking internal mobility within current employers toward infrastructure teams working on AI systems. The timeline for meaningful competency is 6–12 months of focused learning and project work.
LLM Evaluation and Benchmarking
As companies deploy large language models, they need engineers who can design evaluation frameworks, benchmark models against real-world tasks, and identify failure modes. This is an emerging specialisation that sits at the intersection of software engineering, data science, and product thinking. NRIs can build expertise by working with open-source LLM evaluation frameworks, contributing to benchmarking projects, and understanding how to translate business requirements into measurable model performance criteria. This pathway is accessible to engineers without deep machine-learning backgrounds.
MLOps and Model Deployment
Machine-learning operations—the practice of deploying, monitoring, and maintaining ML models in production—is a growing field that bridges software engineering and data science. NRIs with DevOps or platform-engineering backgrounds can transition into MLOps by learning model serving frameworks, experiment tracking, and the operational patterns specific to ML systems. This is a more accessible entry point to the AI infrastructure world than pure systems engineering, and it remains in strong demand.
Security and Compliance in AI Systems
As AI systems proliferate, security concerns—prompt injection, model poisoning, data leakage, regulatory compliance—are becoming critical. NRIs with security backgrounds can specialise in AI security, which combines traditional security expertise with understanding of machine-learning systems. This is a high-value specialisation with relatively low competition.
Alternative Visa Pathways: O-1 and EB-1A for Senior Practitioners
For NRIs who have reached senior levels in their careers, or who have made significant contributions to their field, alternative visa categories may offer more flexibility than H-1B sponsorship.
The O-1 visa is designed for individuals with extraordinary ability in their field. For senior engineers, researchers, and architects in AI and related specialisms, the O-1 can be a viable path. The advantage is that it is not subject to the annual cap or lottery that affects H-1B visas, and it does not carry the same wage-floor requirements. The disadvantage is that it requires demonstrating extraordinary ability—typically through publications, patents, awards, or significant industry recognition. For NRIs who have built strong track records in AI infrastructure, security, or applied research, the O-1 is worth exploring with an immigration attorney.
The EB-1A employment-based green card is another option for individuals with extraordinary ability. Unlike the O-1, which is temporary, the EB-1A is a path to permanent residency. The criteria are similar to the O-1, but the application process is more rigorous. For senior NRIs in high-demand specialisms, particularly those with publications, patents, or significant industry contributions, the EB-1A can be a strategic long-term move. The advantage is that it bypasses the country-based quotas that affect other employment-based green cards, which is particularly valuable for Indian nationals, who face substantial backlogs in other categories.
Both pathways require documentation and legal support, but for NRIs with senior credentials, they offer more control over their immigration trajectory than relying on H-1B sponsorship in an increasingly restrictive environment.
What This Means for NRIs Planning US Tech Careers: Strategic Considerations
For NRIs currently abroad and considering a move to the US tech sector, the calculus has shifted. The traditional pathway—secure an H-1B sponsorship from a mid-sized tech company, work for 5–7 years, and build toward a green card—remains possible, but it is now more competitive and more expensive for employers. This means that the bar for sponsorship is higher: employers will prioritise candidates with specialised skills or senior experience over generalist engineers.
For NRIs already in the US on H-1B visas, the priority is strategic upskilling and internal mobility. Rather than assuming that tenure and performance will lead to promotion and green-card sponsorship, it is prudent to actively move toward specialisms where demand remains strong. This may involve lateral moves within the current employer, or it may require changing employers to access teams working on AI infrastructure, security, or applied research.
For NRIs in mid-career roles facing uncertain sponsorship prospects, exploring alternative visa categories—particularly the O-1 or EB-1A—is worth discussing with an immigration attorney. These pathways are more accessible than they were five years ago, particularly for engineers with publications, patents, or significant technical contributions.
Broader Implications: The Restructuring of the Tech Labour Market
The convergence of AI capex reallocation and stricter H-1B policies is not a temporary disruption; it reflects a structural shift in how technology companies are building and scaling their operations. The era of "add engineers to scale" is giving way to "add GPUs and AI agents to scale." This shift will likely persist regardless of short-term policy changes, because it is driven by fundamental economics: AI infrastructure provides leverage that headcount does not.
For the broader tech labour market, this suggests a bifurcation: strong demand for specialists in AI infrastructure, security, and applied research; structural headcount pressure on generalist roles; and a shrinking middle of mid-level engineers in traditional software development. For the Indian diaspora, which has historically filled that middle tier, the implication is that career advancement now requires specialisation and continuous upskilling rather than tenure and seniority alone.
The policy environment is also unlikely to become more permissive toward H-1B hiring. Even if specific fee structures or wage rules change, the underlying political pressure to limit foreign hiring in tech is durable. For NRIs, this means that relying on H-1B sponsorship as a long-term immigration strategy is increasingly risky. Building toward alternative visa categories, or toward senior roles where sponsorship remains economically rational, is a more prudent long-term strategy.
FAQs: Common Questions for NRIs Navigating the Shifting Tech Labour Market
Should I pursue an H-1B sponsorship now, given the policy changes and layoffs?
The answer depends on your specialisation and career stage. If you have expertise in AI infrastructure, security, or applied research, sponsorship remains accessible and worthwhile. If you are a mid-level generalist engineer, the economics are less favourable for employers, and you may face longer timelines or rejections. If you are considering the move, prioritise companies with strong AI infrastructure investments and roles that align with those investments. Smaller companies and those with less AI capex are less likely to sponsor.
What is the realistic timeline for a green card if I secure an H-1B sponsorship now?
This depends heavily on your country of birth. For Indian nationals, the employment-based green-card backlog is substantial, and timelines can extend to 10+ years in some categories. For this reason, exploring alternative pathways—particularly the EB-1A if you have qualifying credentials—is increasingly important. An immigration attorney can provide a more precise estimate based on your specific situation and the sponsoring employer's category.
If I am already in the US on H-1B, what should my upskilling priorities be?
Focus on specialisms where demand is acute and visa sponsorship remains economically rational: AI infrastructure, security, or applied research. Seek internal mobility toward teams working on these areas, or be prepared to change employers if necessary. Simultaneously, document your technical contributions—publications, patents, significant projects—that could support an O-1 or EB-1A application. This provides optionality if H-1B renewal becomes difficult.
Are there visa categories other than H-1B that might be more accessible?
Yes. For senior engineers and researchers with extraordinary ability or significant contributions, the O-1 visa is not subject to annual caps and does not carry the same wage-floor requirements as the H-1B. The EB-1A green card is another option for individuals with extraordinary ability. Both require documentation of your credentials and accomplishments, but they offer more control over your immigration trajectory. Consult an immigration attorney to assess your eligibility.
Is it still worth pursuing a tech career in the US as an NRI, given these headwinds?
Yes, but with caveats. The US tech sector remains the global centre for innovation and offers unmatched compensation, particularly in specialised areas. However, the pathway is now more competitive and requires strategic specialisation rather than relying on generalist skills and tenure. If you have expertise in high-demand areas—AI infrastructure, security, applied research—the US remains an attractive destination. If you are a mid-level generalist, you may find better opportunities in other markets or in India's growing tech sector. The key is to be realistic about your specialisation and to plan your upskilling and visa strategy accordingly.
Sources: USCIS guidance on H-1B visa requirements and prevailing-wage rules; Federal Reserve statements on technology sector capital expenditure; Bureau of Labor Statistics data on technology employment trends; immigration attorney guidance on O-1 and EB-1A visa categories; industry analyst reports on AI infrastructure investment and technology sector employment patterns.
