I Stopped Applying to “Software Engineer” Roles. Here’s What I’d Target Instead.

Part 1 ended with a claim.

The front door got smaller. But the house is still hiring — just not for what you think.

This is the part where I show you exactly what that means.

Four paths. Real numbers. What you actually need to start each one — not the polished version, the real version. And free resources that aren't just "check out Coursera."

If you're coming in from Part 1, you already know why generic "learn to code" advice is failing people right now. If you're starting here — go read that first. It'll make this one land harder.

Path 1 — Cybersecurity

3.4 million unfilled positions globally. Right now. Not projected. Now.

That number has been climbing for years. And AI is making it worse, not better. Every company expanding its AI footprint is also expanding its attack surface. More AI means more risk means more people needed to manage it. The demand is structural. It doesn't go away when the hype cycle moves on.

What it pays

Entry level: $65,000 to $85,000. Median across all experience levels: $124,910. Senior roles: $150,000 and beyond. Specialize in cloud security or AI security and you're at the top of that range. 53% of employers are actively raising starting pay because they cannot find enough people.

What you actually need

Networking fundamentals. How data moves. What TCP/IP is. How DNS works. What a firewall actually does. That's the baseline. After that — one certification. CompTIA Security+ is the standard entry point. Not glamorous. Just what hiring managers look for.

More important than the cert: proof. Cybersecurity hiring has shifted hard toward demonstrated skill. Labs. Write-ups. Documented problem-solving. A write-up of a CTF (Capture the Flag) challenge on GitHub does more than a certificate sitting in your LinkedIn skills section that nobody looks at.

Realistic timeline

6 to 9 months of focused daily learning to be job-ready at entry level. Not a weekend. Not 30 days. Anyone telling you otherwise is selling something.

Where to start — free

• TryHackMe — hands-on labs, gamified, genuinely good. This is where you build the practical skills employers actually want to see
• Google Cybersecurity Certificate on Coursera — structured, beginner-friendly, no prior experience needed
• CISA's free Cloud Computing Security course — two and a half hours, covers the security-cloud overlap where most of the hiring is concentrated right now

Picking a specialization early — cloud security, penetration testing, identity management, DevSecOps — gives you way more leverage than staying general. Generalists still find work. Specialists find better work faster and don't compete with everyone.

Path 2 — Cloud Engineering

94% of enterprises use cloud services. Global public cloud spending this year: projected past $700 billion. 64% of organizations report cloud skills gaps so severe they're struggling to staff their own infrastructure.

The AI buildout runs on cloud. Every model being trained, every agent being deployed, every API being called — it all lives somewhere. Someone has to build and maintain that somewhere. That someone is a cloud engineer.

What it pays

US average: $151,000. Senior level: up to $183,000. Remote roles — and there are a lot of them — average $135,000. 74% of cloud engineers say they're satisfied with their salaries. That's a high number for any field.

What you actually need

Start with one platform. Not all three. Pick AWS — most widely hired for, best documented, most job postings. Learn it properly before you look at Azure or GCP.

Skills in almost every posting: AWS services, APIs, DevOps basics. Kubernetes pushes you into higher salary brackets once you have the foundation.

Certifications matter more here than in cybersecurity. AWS Certified Cloud Practitioner is your entry point. AWS certs add 25.9% to salary on average and appear in 80% of job postings. One cert, one platform, two real projects. That's your starting stack.

Realistic timeline

6 to 12 months. More to learn than cybersecurity at the entry level but the path is well documented. AWS, Azure, and GCP all have free tier accounts — you can build real infrastructure without spending a dollar.

Where to start — free

• AWS Free Tier — create an account and start building actual things. This is the most important step. Reading about cloud without touching it is useless.
• Great Learning Academy — free cloud computing courses, hands-on projects, covers AWS/Azure/GCP fundamentals
• AWS's own training — platform-specific, practical, free at the foundational level

The highest-paid specializations right now are cloud security engineering, FinOps (cloud cost optimization), and multi-cloud architecture. Knowing your direction early means your portfolio builds toward something specific instead of being a pile of random tutorials.

Path 3 — Data Engineering and AI/ML

AI/ML roles appear in 89% of job postings right now.

Average AI engineer salary in 2026: $206,000. A $50,000 jump from last year. The World Economic Forum says AI has already created 1.3 million new roles globally. BLS projects 34% growth for data scientists through 2034.

Here's why data engineers specifically are impossible to hire fast enough: AI needs data pipelines to function. Before any model gets trained, before any AI product ships, someone has to build the infrastructure that feeds it. That's data engineering. The plumbing of the AI era. Unglamorous. Essential. Extremely well paid.

What it pays

Mid-level data engineers: $119,000 to $149,000 nationally. In major tech hubs: $148,000 to $186,000. Entry-level ML engineers average $95,000. Senior ML engineers: $180,000 to $280,000. Median across all AI/ML engineering positions: $187,500.

What you actually need

Python. That's the real entry point. If you don't know Python, start there — everything in this lane requires it.

After that: SQL for data handling, Apache Airflow for pipelines, Snowflake or BigQuery for cloud data warehousing. For the ML side — you don't need to train models from scratch. Most hiring right now is for people who can work with pretrained models and LLMs. That shift matters because it means the barrier to entry is lower than most people assume.

Realistic timeline

9 to 12 months to be genuinely competitive at entry level. This is the most technical of the four paths. Shortcuts show up fast in interviews here.

Where to start — free

• Kaggle — real datasets, notebooks, competitions. A Kaggle portfolio is one of the most recognized signals for data hiring managers. Build here first.
• 365 Data Science — structured curriculum from zero, free tier available
• fast.ai — practical deep learning, free, built for people who want to understand how things actually work not just run someone else's code

Steepest learning curve of the four. Also the highest ceiling. If you're technically inclined and willing to stay consistent for a year — no other path on this list offers the same salary trajectory over time.

Path 4 — AI Implementation

This is the one nobody's talking about properly yet.

70% of AI projects fail to move past the pilot stage. Not because the technology doesn't work. Because companies can't integrate it into how they actually operate. The gap between "we have an AI tool" and "it works inside our workflow" is enormous — and somebody has to close it.

That somebody is an AI implementation specialist. Take AI that exists, make it function inside a real organization. Wire agents into existing systems. Manage adoption. Bridge the gap between what the vendor promised and what operations actually needs. Healthcare, financial services, manufacturing, and government are hiring hardest for this right now.

This role barely existed two years ago.

What it pays

Entry level: $75,000 to $103,000. Mid-level: $114,000 to $231,000. NVIDIA posted an AI implementation role this year with a base of $184,000 to $287,500. The ceiling is as high as anything else on this list.

What you actually need

Both sides — technical and human. Enough AI literacy to understand how LLMs, agents, and APIs work without necessarily building them. Project or change management capability. And the ability to communicate between engineers and people who don't know what an API is.

Someone from a non-tech background — finance, healthcare, operations — who has learned AI tools seriously is actually competitive faster here than in the other three paths. Domain knowledge plus AI literacy is a combination that's rare and genuinely valued. Companies aren't hiring generalists for this. They want someone who understands a specific industry AND knows how to implement AI inside it.

Realistic timeline

4 to 8 months if you already have a domain background. That combination is the shortcut.

Where to start — free

• Anthropic, OpenAI, and Google documentation — all free, all practical. Start here to understand how the tools actually work
• Build something — connect an AI API to a real workflow. Even a simple automation. That project is your portfolio for this path.
• LinkedIn Learning — AI implementation and change management courses, free with most LinkedIn Premium trials

The job description for this role is still being written industry-wide. Getting in now means you help define what it becomes instead of competing for a slot someone else already shaped.

The One Thing All Four Have in Common

Specialists over generalists. Proof over credentials. People who build things over people who finish courses.

You can start any of these today. Free. With what's listed above.

The question isn't which path is best. It's which one you'll actually stay with for 9 months when it gets boring and hard and you're not sure it's working yet.

That's the real filter. Not talent. Not background. Not whether you studied CS.

Pick one. Start today. Build something real in the first 30 days — something you can point to. That one thing separates you from the majority of people who start and disappear.

The door is smaller. It's still open.

Part 2 of the series: So You Want to Work in Tech in 2026.

Missed Part 1? → Should You Still Get Into Tech? Here's the Honest Answer Nobody's Giving You

Next up: The exact week-by-week learning schedule I'd follow starting from zero — for each one of these paths.

Follow if you want the full series. Find me on LinkedIn for the stuff between posts.