事實脈絡
解讀與影響
从工程侧看，RingCentral 将 Codex 用于 AI 产品的开发流程中，借助代码生成与自动化能力缩短从设计到交付的周期。与此同时，ChatGPT Work 被部署在跨团队协作场景中，帮助工程师与运营人员在同一工作流内完成信息查询、任务分配与进度同步。原文未提供具体的效率提升数据或部署规模细节，但明确指向两个目标：加快 AI 功能迭代，以及减少工程与运营之间的信息断层。
运营侧的整合同样值得关注。RingCentral 试图将原本分散在多个系统中的运营数据集中到统一界面下，使团队能够基于同一套信息做出判断。这种「运营智能」的集中化，意味着从事件响应到资源调度的流程都可能被重新组织。原文未披露具体的技术架构或与现有系统的集成方式，但强调 ChatGPT Work 在这一过程中承担了信息汇聚与分发的作用。
结合同期 arXiv 上关于多智能体系统与 LLM 应用的研究动态来看，企业级 AI 部署正从单一模型调用走向更复杂的协作模式。例如有研究探讨了多 LLM 智能体在目标不一致时的动态治理问题，也有工作关注如何在有限算力下模拟大规模智能体社会。这些方向与 RingCentral 将 AI 嵌入工程与运营全流程的做法形成呼应：当 AI 从工具变为协作节点，如何管理其行为与信息流就变得关键。不过，OpenAI 的案例原文并未提及 RingCentral 是否采用了多智能体架构，相关素材仅作为行业背景参考。
參考來源
來源原文
August 12, 2026
How RingCentral builds AI-native work from engineering to ops
With ChatGPT Work and Codex, RingCentral builds AI product features faster and centralizes operational intelligence.
Contact sales 图片 Company size: Enterprise Region: North America Industry: Technology Products: ChatGPT, Codex Loading… Share With nearly three decades of innovation in business communications, RingCentral has grown into a global company generating more than $2.6 billion in annual revenue, with thousands of employees worldwide. Today, the company is extending its tradition of innovation by embracing AI-native ways of working. By giving every employee room to experiment with ChatGPT Work and Codex, RingCentral has ensured that anyone at the company, regardless of engineering experience, can build transformative products and infrastructure.
“When you put real AI tools in everyone’s hands, the whole company becomes a product organization. Every one of our products—including but not limited to our Agentic Voice AI portfolio of AIR, AVA, ACE—gets sharper as we compress the distance between an idea and a shipped feature, and that’s exactly what AI-native development lets us do.” —Kira Makagon, President & Chief Operating Officer, RingCentral
The AI-Native Challenge
To encourage AI fluency across a global engineering organization, RingCentral’s Office of the CEO sponsored an AI-Native Challenge. Every participant was given ChatGPT Work and Codex, and asked to build a complete, end-to-end project—with no mandated workflow or other constraints.
More than a coding exercise, the challenge immersed employees in the full AI-native development lifecycle, from planning and implementation to testing, documentation, CI/CD, and iteration. Nearly every participant created a working repository, and thousands of employees, including non-technical staff and even executives, delivered functioning projects.
For RingCentral, the challenge is a working model of a broader strategy: using AI internally to build products faster for their customers. The company applies the same Codex-enabled approach to accelerate development of its own AI-powered product portfolio—RingCentral AI Receptionist (AIR), AI Virtual Assistant (AVA), and AI Conversation Expert (ACE)—shortening the distance between an idea and a shipped customer feature.
“The clearest lesson from the challenge was that AI-native development isn’t about replacing engineers—it’s about amplifying them. AI accelerates the entire development cycle, while humans remain in the loop, guiding product requirements, providing business context, making architectural decisions, and ensuring every outcome is tested and verified.”
—Engineering leader at RingCentral who spearheaded the project
Running daily PMO operations with ChatGPT Work
Encouraged by initiatives like the AI-Native Challenge, non-engineering departments at RingCentral have adopted AI-native ways of working. The Program Management Office (PMO) has used ChatGPT Work to build what amounts to an operating system for program management, replacing scattered notes and chat history with AI-powered workflows for status tracking, reporting, release governance, and knowledge transfer.
“ChatGPT brings my project context together. With ChatGPT Work, I can turn that context into actions and execution.” —Vaneet Seth, Senior Manager, R&D Efficiency, PMO, RingCentral One application is automated status reporting: Using ChatGPT Work, the PMO team built workflows that generate notifications from issues tracked across Jira, Google Sheets, CRM systems, and other sources. It’s the difference between walking into a meeting asking what changed and walking in with blockers, owners, and actions already defined.
What started as an open invitation to experiment, with thousands of engineers building from scratch, has matured into the operational backbone of how teams like the PMO run their programs. By reducing manual coordination, ChatGPT Work enables RingCentral’s PMO to handle more projects with greater accuracy.
Across engineering and operations alike, the same pattern holds: giving employees room to experiment with AI doesn’t just build individual skills, it builds the infrastructure the company runs on.
Join the new era of work
More than 1 million businesses around the world are achieving meaningful results with OpenAI. Contact sales
Keep reading
图片 [ How enterprises put AI to work Company Aug 12, 2026
](/index/how-enterprises-put-ai-to-work/) 图片 [ Testing ads in ChatGPT Company Aug 11, 2026
](/index/testing-ads-in-chatgpt/) 图片 [ Daybreak models are now available on AWS Product Aug 11, 2026
](/index/daybreak-models-are-now-available-on-aws/)