RingCentral is betting that the fastest way to build AI products is to let everyone in the company build them. The $2.6 billion business communications giant recently ran an internal ‘AI-Native Challenge’ sponsored by its Office of the CEO, handing every participant access to ChatGPT Work and Codex with a simple ask: build a complete, end-to-end project. No mandated workflow, no constraints.
The results were telling. Nearly every participant produced a working repository, and thousands of employees — including non-technical staff and executives with zero engineering background — shipped functioning projects. That’s not a typo. Executives building software.
One engineering leader who spearheaded the initiative framed the philosophy bluntly: ‘AI-native development isn’t about replacing engineers—it’s about amplifying them.’ Humans stay in the loop for product requirements, business context, and architectural decisions. AI compresses the distance between an idea and a shipped feature. The company is now applying that same Codex-enabled approach to its Agentic Voice AI portfolio — products like AIR (AI Receptionist), AVA (AI Virtual Assistant), and ACE (AI Conversation Expert).
What’s interesting is where this experiment went next. The Program Management Office, a decidedly non-engineering department, used ChatGPT Work to build what amounts to an internal operating system for program management. Instead of scattered notes and chat logs, they now run AI-powered workflows for status tracking, release governance, and knowledge transfer. One application pulls notifications from Jira, Google Sheets, and CRM systems to generate automated status reports — so PMO staff walk into meetings with blockers and owners already defined rather than asking what changed. The pattern is clear: when you give employees room to experiment, they don’t just learn skills. They build the infrastructure the company ends up running on.
