Dr.-Ing. Amir Mabhout

I spent thirteen years in engineering research, got a PhD in chip design, and learned to see the world as interconnected systems. After defending my doctorate, it hit me: I'd been working at the bottom of the technology stack, etching silicon at nano resolution to meet requirements I never questioned. All those specs came from the top, from applications where real people use technology. I realized I wanted to be up there, building things people actually interact with and shape the society, not just implementing someone else's vision in hardware.

In 2021, I found my passion at the intersection of AI, data, and crypto. It wasn't abandoning hardware engineering, it was evolving from it. The same principles that govern electron flow through silicon now help me architect data flows through prompt templates, memories and networks. Electrons and data aren't that different when you think about it.

Here's the truth: I hated coding throughout my entire PhD. Ran from it completely. But when I needed to build applications to realize my vision, programming became joyful. Each bug made me happy to not design chip anymore, since when my design fails, I don't waste expensive silicon wafers and half a year till next chip tape-out shuttle.

Today, I use Claude Code and other AI tools to do most the ground work. I spend good time upfront planning the architecture, mapping data flows, optimizing architecture. What once took months before Claude 4.1 Opus now takes hours at my desk.

I work with LLMs and AI Agent frameworks now, but I'm still that engineer who gets excited about elegant solutions. Whether routing and manipulating electrons through silicon or tokens through inference, the goal is the same: make complex things simple, make slow things fast, and build things that work. Still learning, still failing, still building.