I recently delivered an academic seminar on Spec-Driven Development with AI at the Holon Institute of Technology (HIT). We explored how AI is transforming software engineering, shifting our focus from writing code to defining specifications, controlling AI agents, and orchestrating development.
The photo at the top of this post was not taken at the Holon Institute of Technology (HIT). I chose it because it reflects the excitement I feel these days as I deliver seminars and teach academic courses on Spec-Driven Development. After more than 25 years of teaching in academia, it is an extraordinary feeling to introduce a topic still evolving at the frontier of software engineering.
Seminar Slides: Spec-Driven Development with AI
The seminar slides are available online on Speaker Deck. The presentation explores the shift from traditional code-first development and vibe coding to Spec-Driven Development, where specifications form the foundation of AI-assisted software engineering.
Using Kiro IDE as a practical example, I covered the development process, steering documents, requirements, technical design, agent hooks, subagents, and orchestration patterns. We also explored parallel task execution, automated workflows, cloud-based AI agents, and best practices for maintaining control over increasingly autonomous development processes.
Teaching at the Frontier of Software Engineering
After more than 25 years of teaching advanced software development topics in academia, I found this seminar particularly exciting. Throughout my academic career, I have taught many advanced technologies, programming languages, and software engineering concepts. However, this seminar felt different. For the first time, I felt I was introducing a topic at the very frontier of software engineering, as the field itself undergoes a fundamental transformation. Exploring this emerging paradigm with an academic audience, while its practices and tools are still evolving, made the experience especially meaningful. It reminded me why, after all these years, I remain so passionate about teaching, learning, and software development.
Software engineering is entering an exciting new era. As AI takes on more coding responsibilities, our ability to define intent, design systems, and orchestrate agents becomes increasingly important. I’m excited to explore and teach this transformation.







