About

I'm Sertan. My work combines engineering leadership with hands-on building.

My first research was on augmented reality and computer vision in the mid-2000s. The ideas were right, the timing was early, and I chose an engineering career over completing the PhD. More than fifteen years of building and running distributed systems followed, across automotive, online retail, location intelligence and embedded finance, where I currently lead engineering. In 2024 I closed the loop and completed an MSc in Artificial Intelligence at the University of Bath, with distinction.

My focus now is the practical adoption of AI in software engineering: agentic development workflows, Model Context Protocol (MCP) servers, spec-driven development, and the leadership decisions that determine whether any of it actually pays off. In my own work, I use a conservative autonomy model in which deterministic systems own the loop, the large language model (LLM) advises, and a safety policy validates every action.

Much of my recent building is done by coding and review agents working within boundaries I define. I set the architecture and specifications, decide what must remain deterministic, choose the review lenses, and judge the evidence before anything merges. I remain accountable for the resulting system, regardless of how many lines I type myself.

I write about these experiments because building and operating the systems gives me a practical way to examine AI in engineering without relying on either the hype or the doom around it.

The AI and the Industrial Revolution series is a good place to start and is best read in order.