Eno Reyes of Factory on AI agents for enterprise code modernization: $80M SI quotes being disrupted

Summary

Factory, co-founded by Eno Reyes (CTO), is targeting a specific and quantifiable pain point in enterprise software: large-scale code modernization projects that systems integrators currently price at $80 million or more. The company's platform deploys AI agents across the full software development lifecycle, covering coding, code review, incident response, and documentation, with a focus on organizations running more than 1,000 engineers.

The Beachhead

Factory's clearest use case is migration and modernization at scale. The canonical example Reyes cites involves migrating 185 codebases to a new framework across a team of 3,000 developers over six months. An SI quoted that project at $80 million. Factory's pitch is that agent-driven automation can displace a significant portion of that spend.

The legacy infrastructure opportunity is substantial. Reyes points to COBOL-based financial systems, where 5 to 10% of global payment transactions still run on code written 40 years ago, with no remaining engineers who know the stack. These systems represent existential technical debt for their owners, and the addressable market for modernization runs to hundreds of billions of dollars by Reyes's framing.

The AI Productivity Gap

Factory only deploys into companies that have already adopted an AI-native IDE or autocomplete tooling. That qualifier surfaces a revealing pattern: enterprises are being promised 5x to 10x developer productivity gains from AI tools but are seeing only 10 to 15% improvements in practice. Reyes attributes the gap to workflow inertia. Accelerating individual developers at the margin produces modest gains, while parallelizing and automating work at the pipeline level is where the larger productivity shift occurs.

Market Structure Outlook

Reyes expects the enterprise coding agent market to follow an 80/20 split: a broad platform layer serving the majority of large enterprises with common needs around refactoring, migration, and modernization, alongside a smaller tier of specialized providers focused on narrow verticals or specific languages. He does not see the market consolidating into a pure oligopoly or fragmenting entirely, but rather stratifying in a pattern similar to the cloud infrastructure market, with hyperscalers dominant and a long tail of niche players capturing the remainder.