Michael Petychakis is Chief Technology Officer at Dialectica and a technologist whose work spans robotics, APIs, AI and engineering leadership. His path took him from electrical and computer engineering at the National Technical University of Athens to a social humanoid robotics lab in the UAE, and then to PhD research on APIs and the web. Before joining Dialectica, he served as CTO at ORFIUM, building products and engineering teams for the global music industry. He writes at APILama and mentors startups and aspiring founders. His recurring question is how to build software that people can understand, trust and change.
Michael Petychakis is Chief Technology Officer at Dialectica and a technologist whose work spans robotics, APIs, AI and engineering leadership. His path took him from electrical and computer engineering at the National Technical University of Athens to a social humanoid robotics lab in the UAE, and then to PhD research on APIs and the web. Before joining Dialectica, he served as CTO at ORFIUM, building products and engineering teams for the global music industry. He writes at APILama and mentors startups and aspiring founders. His recurring question is how to build software that people can understand, trust and change.
Once upon a time, three brothers set out with a map, three gifts, and a perfectly reasonable plan. The world had other ideas.
In 2015, my PhD research explored how software agents could discover actions and navigate unfamiliar services. I imagined an agentic web. I was rather less accurate about the engine that would power it.
A Hypermedia Fairytale takes that old question on a new journey: through enchanted markets, misleading signs, and doors that make promises. Along the way, we meet the problems hiding inside modern AI systems: understanding what an API means, distinguishing a suggestion from permission, coordinating different capabilities, and deciding what to do when an action may have succeeded but the reply never arrived.
This visual keynote connects software craftsmanship, APIs, and AI agents through a fairytale and a few small API examples. No specialist knowledge of hypermedia or AI is required. Students, engineers, and product leaders will leave with practical questions for designing systems that explain their actions, respect their limits, and recover honestly when things go wrong.
When machines can write the code, what does it mean to build software well?
