Building a Trustworthy Agentic Software Factory
Description
Software development is crossing a threshold. AI agents are moving beyond code generation. They are beginning to plan work, implement changes, test systems, review results and operate complete development workflows.
This changes the central challenge for technology leaders. The question is no longer simply how to generate more code, but how to build a software production system that can operate with increasing autonomy while remaining secure, reliable and accountable.
This talk introduces the agent-centric development cycle, a practical framework that development and platform teams can use to build a trustworthy software factory. It explains how organisations can give agents clear intent and effective constraints, establish independent layers of verification, and progressively increase autonomy as evidence of quality and control grows.
It also examines what this transformation means for developers and the shape of development teams. As agents take on more implementation work, developers increasingly define intent, design systems and engineer the environments and verification mechanisms that make autonomous work trustworthy. Humans remain responsible for judgment and accountable for the resulting outcomes.
Attendees will leave with a practical model for moving beyond isolated coding assistants toward an AI-enabled software factory that delivers greater speed and scale without sacrificing enterprise quality, security or control.
