An AI software factory is a managed system for turning defined requirements into validated software changes using AI agents, connected tools, automated checks, and human oversight. Unlike a coding agent that performs development work, a software factory connects business requests to context, tools, people, validation, and release processes. At Camplight, three client deployments over nine months revealed that the critical bottleneck is verification, as evals and guardrails remain a moving target. The team generalized their experience into Nest, a public model based on open-source Org Ops infrastructure, visualizing their operating questions without revealing confidential client systems. This generalized model serves as an evaluation framework rather than a universal product checklist, illustrating the operating questions behind successful deployments.

The distinction between a software factory and a coding agent lies in the system around execution. A factory connects intent to relevant context, execution, verification, required approval, release, and feedback loops. Different flow components may use agents, ordinary code, existing pipelines, or people. Critically, the system does not replace CI/CD but must connect agent-generated work to existing build, test, and deployment mechanisms. Leaders must govern the operational questions of intent, context, and feedback to deliver validated software changes effectively.