AI is no longer just a standalone tool, an assistant, we use on demand in software development.

More and more, we are talking about skills, MCPs, hooks, custom agents, shared context, and workflows designed to help us build better and faster. At the same time, open-source solutions such as nWave are starting to appear, giving us a ready-made foundation we can benefit from without reinventing the wheel. But the real challenge is not just which tool to use. It is how to shape a workflow that actually makes sense, and then how to roll it out across a team in a coherent way.

In this session, we will explore which building blocks usually matter most when designing these workflows, what kinds of decisions come up during adoption and their implications, and what challenges appear when you try to bring all of this into real teams. This is a talk to share practical experience, open up discussion, and bring some structure to a space that right now is moving very fast, full of novelty, and, also full of noise.