Most developers already use AI to generate code.
But what changes when AI moves beyond autocomplete and starts participating in the software development lifecycle?
While developing nVibes, AI agents are used alongside GitHub, Scrum and Azure to analyze tasks, inspect repositories, propose implementations, create changes and support software delivery.
This session explores a new development model in which AI is treated less like a coding assistant and more like a temporary member of the engineering team.
We will examine where agents fit into a Scrum-based workflow, what context they require, which tasks they perform well, and where human review remains essential.
During the live demo, we will give an AI agent a real development task from the nVibes backlog and follow the process from requirement to implementation.
The interesting question is no longer:
“Can AI write code?”
It is:
“How do we design a development process when both humans and AI agents can perform engineering work?”
What attendees will learn
- The difference between AI-assisted coding and agentic development.
- How AI agents can work with repositories, issues and development tasks.
- How agent work can fit into GitHub and Scrum workflows.
- Why context, acceptance criteria and architecture become even more important with AI.
- Where human review, governance and accountability remain essential.
Live Demo
Give the agent a genuine small nVibes Issue.
For example:
Add/modify a small feature or fix a controlled bug.

Then:
Issue → Agent analysis → repository → implementation → diff → tests → PR/review