Why I Combine Electrical Engineering With AI Instead of Picking One
Engineering gives you rigor. AI gives you leverage. Here's how I think about combining both instead of specializing too early.
Kamal Benaouija
April 2, 2026
People regularly ask why I don't just "pick a lane" — electrical engineering or software or AI. My answer: the most interesting problems live at the intersection, and the intersection is exactly where fewer people are equipped to work.
Engineering Gives You Rigor
Electrical engineering trains a very specific kind of thinking: systems have constraints, tolerances, and failure modes that don't care about your intentions. A circuit either meets spec or it doesn't. That discipline transfers directly into how I approach software and AI systems — I default to asking "what are the constraints and failure modes here?" before I ask "what's the fastest way to ship this?"
AI Gives You Leverage
Where engineering teaches rigor, AI — specifically LLMs and automation — teaches leverage. A well-designed prompt pipeline or automation script can replace hours of manual, repetitive work. Applied correctly, AI doesn't replace engineering judgment; it multiplies it.
Where They Meet
Here's a simplified mental model I use when scoping any new project:
| Layer | Question | Tool |
|---|---|---|
| Constraints | What physically/technically must be true? | Engineering fundamentals |
| Automation | What repetitive work can be removed? | AI / scripting |
| Interface | How does a human interact with this? | Software / UX |
Most people are strong in one layer and weak in the other two. Being competent across all three is rare — and it's exactly the skill set that lets you build a complete product instead of a single component of one.
A Concrete Example
On a recent automation panel project, I:
- Designed the circuit and control logic (engineering layer)
- Wrote firmware to expose sensor data over a simple protocol (automation layer)
- Built a real-time dashboard so the data was actually usable (interface layer)
No single layer would have produced something useful on its own. The value was in connecting all three.
The Takeaway
I don't think of myself as "an electrical engineer who also codes" or "a developer who also knows hardware." I think of myself as someone who treats hardware, software and AI as one system — because in the real world, that's exactly what they are.