The AI-Native Challenge: Bottom-Up Innovation

RingCentral’s approach to AI adoption centers on an internal 'AI-Native Challenge' sponsored by the Office of the CEO. Rather than mandating specific workflows, the company provided employees with ChatGPT Work and Codex, tasking them with building end-to-end projects. This initiative served two purposes: it democratized technical development across the company and acted as a testing ground for accelerating the development of their own AI-powered product portfolio (AIR, AVA, and ACE).

Key takeaways from this experiment include:

  • Amplification, Not Replacement: The primary insight is that AI accelerates the development lifecycle—from planning and coding to documentation and CI/CD—while humans remain essential for architectural decisions, business context, and rigorous verification.
  • Cross-Functional Participation: By removing barriers to entry, thousands of employees, including non-technical staff and executives, successfully created functioning repositories, proving that AI tools can bridge the gap between an idea and a shipped feature.

Operationalizing AI in the PMO

Beyond engineering, RingCentral applied AI-native principles to the Program Management Office (PMO) to replace fragmented documentation and manual status tracking. They built an AI-powered operating system that aggregates data from Jira, Google Sheets, and CRM systems to automate reporting.

This shift provides several operational advantages:

  • Contextual Awareness: Instead of manual status updates, the PMO now uses AI to generate notifications and identify blockers, owners, and required actions automatically.
  • Scalability: By reducing the manual coordination burden, the PMO can manage a higher volume of projects with increased accuracy, turning scattered chat history and notes into a centralized, actionable knowledge base.