At enterprise scale, engineering work takes place in long-lived, interconnected codebases shaped by years of development. EY needed a solution that could:

• Operate inside existing systems engineers already rely on (GitHub, Jira, Slack, internal DevOps tooling)
• Handle execution-heavy work such as refactoring, code changes, and documentation updates across large repositories
• Preserve enterprise standards and institutional knowledge
• Meet enterprise requirements for quality and security

Scaling execution inside existing engineering workflows

For agents to operate in production and daily workflows, they must operate inside the systems engineers already rely on. At EY, Droids run directly within existing development environments. This allowed Factory to quickly take on execution-heavy work while maintaining established review processes.

In this revised workflow:

• Droids execute defined engineering tasks
• Engineers retain architectural ownership
• Engineers maintain final approval before merge

This enables true delegation without introducing parallel workflows or bypassing established review processes.

Teaching agents enterprise memory

Droids are designed to operate in enterprise environments by drawing on a shared knowledge graph that captures organizational context and historical codebase structure. This allows Droids to implement complex changes across repositories while preserving enterprise standards and institutional knowledge.

"We're making the Droids available firm-wide because they have become a true cornerstone of scalable software strategy. The Factory team took our partnership even further and made sure we can harvest the data we need to track impact with our custom-built performance index tool," says Martin Nicolas Ausilio.

Operationalizing at global scale

The initiative is led by Stephen Newman, EY's Global Head of Engineering, and Martin Nicolas Ausilio, Director of Engineering. Supported by EY's Customer Engineering organization, the deployment will continue expanding across additional business units and regions throughout 2026.

EY's deployment demonstrates how agent-native development can be adopted at enterprise scale. By integrating agents directly into production workflows and holding them to the same standards as engineers, EY has established a repeatable model for adoption that maintains engineering quality while materially increasing velocity.

The practices EY has put in place now serve as a reference point for other enterprises evaluating how to introduce agent-native development across their engineering organizations.

Looking to accelerate your software development? Learn more about Factory.