The tech job market everyone prepared for no longer exists.

If you’ve been grinding LeetCode problems and polishing your algorithm skills, hoping to land that software engineering role—you might want to sit down for this. According to Indeed’s 2025-2026 hiring data, software engineering job postings have dropped to 65% of their February 2020 levels. That’s not a dip. That’s a fundamental restructuring.

But here’s what the doom-and-gloom headlines miss: this isn’t the death of tech jobs. It’s a massive reallocation. While traditional coding roles contract, infrastructure, security, and AI-adjacent positions are growing faster than companies can fill them. The question isn’t whether there are jobs—it’s whether you’re positioned for the jobs that actually exist.

The Numbers That Should Change Your Career Plan

Let’s be specific about what’s happening, because vague anxiety helps nobody.

Robert Half’s 2026 analysis found that traditional software development roles now represent only 10% of in-demand tech positions. Meanwhile, cloud engineering, data analytics, and AI/ML roles have surged to the top of hiring priorities.

The Bureau of Labor Statistics projects computer and IT occupations will grow faster than average through 2034, with roughly 317,700 openings annually. But the distribution matters enormously:

Role CategoryProjected Growth (2023-2033)
Information Security Analysts33%
Computer and Information Systems Managers17%
Computer Network Architects13%
Computer Systems Analysts11%
Database Administrators9%

Notice what’s not leading that chart? Traditional software developer roles.

This isn’t speculation. Ravio’s 2025 tech hiring report found that entry-level software engineering positions (P1 and P2 levels) experienced a 73% decrease in hiring rates compared to just 7% across all tech job levels. If you’re a fresh computer science graduate, your unemployment rate is currently 5.8%—higher than the general unemployment rate.

Why This Is Happening (And Why It Won’t Reverse)

The obvious answer is “AI is replacing developers.” That’s partially true, but it misses the bigger picture.

The AI Productivity Paradox

Companies using AI coding assistants like GitHub Copilot report 30% productivity gains while keeping headcounts flat. Roughly 75% of engineers now use AI assistance for routine coding tasks. This doesn’t eliminate developers. It means fewer developers can accomplish more. The AI impact on IT jobs is real, but it’s more nuanced than mass replacement.

What AI can’t easily automate: designing systems, securing infrastructure, managing complex cloud deployments, and making judgment calls about architecture. The tasks that require understanding context, business requirements, and failure modes at scale.

The End of Growth-at-All-Costs

The 2021-2022 hiring boom was an anomaly, not a baseline. Near-zero interest rates let companies hire aggressively without worrying about runway. When rates normalized, so did hiring—just with a painful correction attached.

Amazon, Intel, and Microsoft alone account for roughly 64% of the 211,000+ tech layoffs projected through 2026. These aren’t struggling companies. They’re recalibrating from an unsustainable expansion.

The Infrastructure Debt Comes Due

Here’s the part nobody’s discussing: all those applications built during the hiring boom need to be deployed, secured, scaled, and maintained. That’s infrastructure work. The ratio of developers to infrastructure engineers has been inverted for years, and companies are finally feeling the pain of underfunding platform teams.

If you’ve ever worked somewhere with one overwhelmed sysadmin supporting fifty developers, you’ve seen this dynamic firsthand. That model doesn’t scale, and organizations are correcting by investing in DevOps, platform engineering, and cloud architecture.

Where the Demand Actually Is

Let me be specific about what’s hiring, because “infrastructure” is vague enough to be useless.

Security: The Permanent Talent Crisis

Information security analysts are projected to grow 33% through 2033. That’s not a typo. The cybersecurity talent gap has been a talking point for years, but the math hasn’t changed: organizations are desperate for qualified security professionals, and “qualified” keeps getting redefined upward as threats evolve.

If you’re considering the cybersecurity career path, the runway is long. But be realistic: entry-level security roles often want you to understand systems administration, networking, and at least one scripting language first. The path through help desk to sysadmin to security analyst remains well-worn for good reasons.

The cybersecurity field also benefits from regulatory tailwinds. Compliance requirements around data protection, incident reporting, and security audits keep creating new positions even when companies tighten other budgets. Security is rarely the first team to face cuts—usually it’s one of the last.

Cloud Engineering: Still Growing, Higher Bar

AWS, Azure, and Google Cloud continue expanding, and organizations need people who can manage their presence on these platforms. The cloud certification roadmap is well-documented.

What’s changed: junior cloud roles are harder to find. Companies want engineers who can design architectures, not just follow deployment scripts. The way in is increasingly through infrastructure experience first—understanding networking, Linux systems, and how applications actually run on servers.

For hands-on Linux practice, Shell Samurai offers interactive terminal challenges that build real command-line muscle memory—the kind of fundamentals that distinguish candidates who actually understand infrastructure from those who only know cloud consoles.

AI/ML Operations: The New Frontier

Ravio’s data shows AI/ML roles grew 88% in 2025 compared to the previous year. The share of AI/ML jobs in tech has jumped from 10% to 50% in just two years.

But here’s what that actually means on the ground: companies aren’t hiring armies of machine learning researchers. They’re hiring people who can deploy, monitor, and maintain AI systems in production. MLOps, AI infrastructure, and AI governance roles are exploding. The AI skills that matter are increasingly about integration and operations, not building models from scratch.

Think about it this way: for every data scientist building a model, there are three or four engineers needed to put that model into production, monitor its performance, handle edge cases, and ensure it doesn’t break at scale. That’s where the jobs are. If you understand containers, CI/CD pipelines, and monitoring systems, you’re already halfway to an ML engineering role.

Platform Engineering: DevOps Grows Up

Platform engineering emerged because DevOps in practice often meant “the ops team learns to code” rather than genuine collaboration. Platform teams build internal tools and infrastructure that let developers ship without needing to understand every layer of the stack.

If you’re already in the DevOps space or considering the sysadmin to DevOps transition, platform engineering is a natural evolution. The work is increasingly about creating abstractions and developer experience, not just maintaining pipelines.

What This Means For Your Career Strategy

Let me be direct: if you’re currently pursuing a pure software development path, you have options, but they require honest assessment.

If You’re Entry-Level or Career Switching

The traditional “learn to code, get a dev job” pipeline is congested. That 73% drop in entry-level hiring isn’t just noise—it reflects a genuine reduction in demand for junior developers who need significant mentorship.

Consider these alternative entry points:

Infrastructure-first path: Help desk → system administration → cloud or security specialization. This path has more entry points, clearer progression, and ends up at in-demand roles.

Security-adjacent path: Start with CompTIA A+ and Network+, build a home lab, practice on platforms like TryHackMe and HackTheBox, then pursue Security+.

Data/Analytics path: SQL remains foundational. Data analyst roles often have lower barriers than development roles and position you well for AI-adjacent work.

The key insight here: these paths aren’t consolation prizes. They’re often faster routes to six-figure salaries than the crowded developer pipeline. A senior cloud architect or security engineer earns as much or more than a senior software engineer, with fewer candidates competing for each opening.

If You’re a Mid-Level Developer

You have skills that transfer. The question is whether you want to specialize laterally or vertically.

Lateral moves: Backend developers with infrastructure experience can transition to SRE or platform engineering relatively smoothly. Your understanding of how applications work gives you an advantage over pure-ops candidates.

Vertical specialization: AI tooling, security, and performance engineering are areas where developer skills combine with specialized knowledge. A developer who understands how to build secure, performant systems is more valuable than one who only knows frameworks.

The skills you should add: Regardless of direction, Docker, Kubernetes, Terraform, and at least one major cloud platform are becoming table stakes. If you haven’t touched infrastructure, now is the time. Start small: containerize an application you’ve built, deploy it to a cloud free tier, set up a basic CI/CD pipeline. These aren’t side projects—they’re resume builders that demonstrate you understand the full delivery lifecycle.

If You’re in Infrastructure Already

Congratulations on picking the right horse, even if by accident. Your next move is specialization.

Security specialization: Infrastructure pros who understand security are unicorns. The IT support to cybersecurity path is well-documented, and your operational experience is a genuine advantage.

Cloud architecture: Move from operating cloud resources to designing systems. This means understanding cost optimization, resilience patterns, and cross-service integration deeply. The difference between a cloud engineer and a cloud architect is the ability to look at business requirements and translate them into technical designs that balance performance, cost, and reliability.

Platform engineering: If you enjoy building tools and automation, this specialty combines infrastructure knowledge with product thinking. Your internal customers are developers, and your success is measured by how quickly they can ship reliable software. It’s a different skill set than pure operations, but your infrastructure background gives you credibility and context that’s hard to fake.

The Skills That Actually Matter Now

Beyond specific roles, certain skills matter regardless of which direction you go.

Judgment Over Execution

AI can write code snippets. AI can suggest configurations. What AI cannot do reliably: decide what should be built, evaluate tradeoffs, understand business context, or predict failure modes. The skills that matter increasingly involve making decisions, not just implementing them.

This is why soft skills matter more than ever. The ability to communicate with non-technical stakeholders, understand business requirements, and advocate for technical investments separates senior from junior—regardless of technical domain.

Operational Experience

Running systems in production teaches lessons that tutorials cannot. If you haven’t experienced an outage, debugged a performance problem under pressure, or made a change that broke something important—you’re missing context that hiring managers value.

There’s a reason so many job postings say “experience with production systems required.” It’s not gatekeeping. It’s recognition that production environments have failure modes and complexity that simply don’t exist in development. The person who’s been woken up at 3 AM by a pager approaches system design differently than someone who’s only built things that work on their laptop.

This is why home labs and hands-on practice matter. Platforms like VirtualBox, Proxmox, and cloud free tiers let you build real environments. Shell Samurai provides structured challenges for building command-line proficiency. The goal isn’t credentials—it’s genuine experience managing complexity.

Adaptability Over Specialization (Paradoxically)

The most resilient tech careers involve T-shaped knowledge: deep expertise in one area, broad familiarity across many. Specialists who can only do one thing are vulnerable to that thing becoming obsolete. Generalists who know a little about everything struggle to provide unique value.

The sweet spot: become genuinely good at something in demand (security, cloud architecture, platform engineering), while maintaining enough breadth to pivot when the market shifts again—because it will.

What Not To Do

A few common mistakes worth avoiding:

Don’t panic-pivot: The job market for traditional developers is tighter, not dead. If you love building software and are good at it, the path forward might be finding the right company or niche—not abandoning your expertise entirely.

Don’t chase certifications without skills: Cloud certifications matter, but only when backed by actual ability. The certification market has become saturated enough that credentials alone don’t differentiate. Stop collecting certifications and start building.

Don’t ignore the fundamentals: Every infrastructure role eventually requires understanding networking, operating systems, and how applications actually work. Skipping to the cloud without understanding what’s underneath creates fragile knowledge.

Don’t assume AI will do your job: AI tools augment skilled professionals; they don’t replace them. The engineers who treat AI as a productivity multiplier are pulling ahead of those who either ignore it or assume it will handle everything.

The Bottom Line

The tech job market isn’t dying. It’s evolving. Traditional software development roles are consolidating while infrastructure, security, and AI operations roles expand. This is uncomfortable if you planned a different career, but it’s navigable with honest assessment and strategic adjustment.

The IT job market in 2026 rewards people who can design systems, secure them, deploy them at scale, and make judgment calls about tradeoffs. These aren’t new skills. They’re the skills that infrastructure and operations professionals have always needed. What’s new is that more of the industry recognizes their value.

Whether you’re entering tech, transitioning specialties, or advancing in your current path, the same principles apply: build real skills through hands-on practice, understand systems deeply enough to make good decisions, and stay adaptable enough to shift when the market shifts again.

Because it will. It always does.

Frequently Asked Questions

Is software engineering still a viable career choice in 2026?

Yes, but with caveats. The field is more competitive, entry-level hiring has contracted significantly, and AI tools are changing what individual productivity looks like. Software engineering remains viable if you’re genuinely skilled, focused on areas that resist automation (architecture, complex systems, security-conscious development), and realistic about the current market. The days of “learn JavaScript in 3 months and get a $100K job” are probably over.

Which tech roles are most resistant to AI displacement?

Roles involving judgment, physical systems, and high-stakes decisions tend to be more resistant. Security professionals who assess risk and respond to incidents, infrastructure engineers who design resilient systems, and architects who make tradeoff decisions all require context and judgment that current AI struggles with. Any role that involves understanding why something should be built—not just how—has more staying power.

How long does it take to transition from development to infrastructure?

Typically 6-12 months for a meaningful transition if you’re deliberate about it. The first step is usually learning Linux fundamentals deeply, then adding cloud platform experience, then containerization and orchestration. Your development background helps—understanding how applications work gives you context that pure-ops people often lack. The sysadmin to DevOps path offers a good roadmap that developers can adapt.

Should I get cloud certifications now or wait?

Get them, but with a strategy. Start with one major platform (AWS or Azure) rather than spreading across all three. Pair certification study with actual hands-on work—free tier accounts, personal projects, home labs. Certifications without demonstrated practical skills are increasingly seen as checkbox exercises rather than genuine qualifications.

What’s the outlook for cybersecurity careers specifically?

Extremely strong. The 33% projected growth for information security analysts reflects genuine, sustained demand that shows no signs of slowing. The challenge isn’t finding opportunity—it’s meeting the bar for entry, which typically requires solid IT fundamentals first. The cybersecurity career path is real, but it usually runs through help desk, networking, or system administration before reaching dedicated security roles.