Skip to content

From Drilling Engineer to AI Engineer: Why Domain Expertise Matters in AI

After 22 years in oil & gas drilling operations, I made the leap to AI engineering. Many people ask me why I didn't make the switch earlier—or why I bothered with AI at all when I had a successful engineering career. The answer is simple: I realized that my industrial experience wasn't something to leave behind. It was my competitive advantage.

My Journey: From BASIC to Python to AI

My relationship with code started early. As a kid, I was writing BASIC and assembler on my father's ZX Spectrum. That curiosity never left me.

Through university and my engineering career, I picked up VBasic, Matlab, C, C++, and C#. Each language served a purpose, but none of them clicked the way Python did when I discovered it about 8 years ago. Python became my go-to language—versatile enough for quick scripts, powerful enough for serious applications.

But here's the thing: I was self-taught for years, and while I could write code, I couldn't really build things. That changed when I joined the PDM program at PyBites with Bob Belderbos as my coach. That's when I learned to develop real projects, not just scripts.

The "Aha" Moment

The turning point came when I started exploring automation tools like Make and n8n. I saw how Python could connect with my industrial experience in ways I hadn't imagined. Suddenly, the problems I'd been solving manually for two decades—data processing, workflow optimization, knowledge management—could be automated intelligently.

That's when AI entered the picture. Not as a replacement for my engineering knowledge, but as an amplifier.

Why AI Projects Fail Without Domain Knowledge

I've seen AI projects fail. Not because the technology was wrong, but because the team didn't truly understand the problem domain.

Here's what happens: a company hires brilliant AI engineers who know everything about transformers, embeddings, and fine-tuning. They build technically impressive systems. But when the system meets real operational data, it falls apart. Why? Because they didn't understand:

  • How the data is actually acquired in the field
  • What the numbers really mean in operational context
  • Which edge cases matter and which don't
  • How decisions are actually made on the ground

In industries like oil & gas, manufacturing, or energy, the data is highly specific. Someone unfamiliar with the domain simply cannot extract meaningful insights from it—no matter how sophisticated their AI models are.

What I Bring That Generic AI Engineers Don't

After 22 years in drilling operations across companies like Repsol, ADNOC, and Moeve, I understand:

  • How industrial data is acquired: sensor limitations, manual logging practices, data gaps
  • How to interpret operational metrics: what's normal, what's alarming, what's noise
  • How decisions are made under pressure: the real workflow, not the documented one
  • What knowledge is critical: the procedures that matter vs. the ones nobody reads

When I build a RAG system for technical documentation, I don't just index PDFs. I understand which sections engineers actually need at 3 AM when something goes wrong.

My Advice for Engineers Considering the Transition

If you're an industrial engineer thinking about moving into AI, here's what I'd tell you:

Prepare thoroughly. Learn Python. Understand the fundamentals of machine learning and LLMs. Build projects, not just tutorials.

But don't trust AI blindly. Your domain knowledge and intuition are irreplaceable. AI should be a catalyst—something that amplifies your expertise, not replaces it. The engineers who thrive in AI are those who can look at a model's output and say, "That doesn't make sense in the real world."

Your experience is your edge. You're not starting from zero. You're starting with 10, 15, 20 years of context that no bootcamp can teach.

My Advice for Companies Hiring AI Freelancers

If you're evaluating AI consultants or freelancers, consider this:

Trust freelancers who've walked in your shoes. Someone who has worked in your industry and then transitioned to AI understands your problems at a deeper level. They won't need months to learn your terminology or workflows.

Domain expertise accelerates everything. Requirements gathering is faster. Solutions are more practical. Edge cases are anticipated. You spend less time explaining and more time building.

The Future Belongs to Hybrid Experts

The AI landscape is crowded with technically skilled people who've never worked outside tech. That's not a criticism—it's an observation.

But for complex industrial applications, the real value comes from people who can bridge both worlds. People who understand the technology and the domain. People who can talk to engineers and data scientists equally.

That's the niche I've carved out. And if you're an engineer with decades of experience wondering if it's too late to make the jump—it's not. Your expertise is exactly what the AI industry needs.

Ready to discuss how AI can solve your industrial documentation challenges? Book a free strategy call and let's talk.