企业面临的挑战已不再是证明AI智能体能够工作，而是让它们足够可靠，能够在生产环境中执行高价值任务。智能体的行为还必须随着产品、政策和用户行为的变化而适应。这需要的不仅仅是模型：还需要系统、评估和部署专业知识，以便在这些条件变化时改进智能体，同时不放弃控制权。
我们推出OpenAI Presence，这是一款经过实战检验的产品，可帮助企业部署值得信赖的AI智能体，这些智能体能够回答问题、解决问题、使用公司系统、执行已批准的操作，并在需要时上报给人工。通过与客户在企业级规模上多年的合作经验证明，Presence将模型推理与策略、护栏和上报规则相结合，以验证准确性和性能。
每次部署都从一个特定任务开始，例如解决账单问题、支持保险理赔或解决员工IT服务请求。智能体仅接收该任务所需的知识和系统访问权限。公司设定策略：智能体可以做什么、何时需要批准、以及何时应由人工接管。上线后，生产会话和上报会揭示差距。Codex提出团队可以测试和批准的更新，帮助智能体随着客户行为的变化而适应。
OpenAI与每位客户紧密合作，识别每个高价值工作流程，连接必要的知识和系统，建立权限和策略，测试智能体，并将其投入生产。随着部署的扩展，OpenAI和选定的系统集成商可以继续提供支持。Presence是与OpenAI研究团队紧密协作开发的，每次部署的通用见解都会为持续的研究和产品开发提供信息，并随着时间的推移为所有客户改进产品。
今日起可用于语音和聊天智能体
如今，Presence支持跨语音和聊天的实时体验，例如客户支持、外呼销售和高风险内部工作流程。客户可能用它来解决账单问题——从理解请求和验证客户身份，到查找账户信息、应用公司政策以及执行已批准的操作。
公司决定哪些内容在部署中保持一致——例如策略、评估和上报规则——以及哪些内容应针对每个工作流程或渠道进行更改。这让团队能够基于有效的内容进行构建，并扩展到新的用例，而无需从头开始。
Presence整合了团队在生产中运行智能体所需的组件：策略和标准操作程序、护栏、已批准的操作、模拟、评估工具以及由Codex驱动的改进流程。
这些组件共同帮助团队连接公司系统、定义智能体应如何表现、评估性能、执行策略以及管理上线后的变更。
在领先企业中经过验证
Presence是通过多年与客户部署智能体而构建的，是一款雄心勃勃的产品，其形态由关键任务环境的需求塑造。每次部署都会产生见解，持续为OpenAI的研究和产品开发提供信息，为客户带来累积效益。
Presence为OpenAI的英语电话支持渠道（1-888-GPT‑0090）提供支持，处理开放式请求、验证来电者身份、使用账户上下文并执行已批准的操作。在数周内，它达到或超过了我们用于评估一线人工支持质量的基准，现在无需人工协助即可解决75%的来电问题。与我们的上线团队合作，其由Codex驱动的改进循环在短短10天内将人工交接减少了15个百分点。
领先企业也在基于相同的经过验证的基础进行构建：
- BBVA正在探索为墨西哥的日常银行需求提供AI驱动的语音支持。
- SoftBank正在测试自然的日语客户对话。
- IAG正在探索在恶劣天气等高需求事件期间提供及时支持。
1 / 3
为上线前、中、后的信任而构建
在Presence部署到达用户之前，团队可以针对常见请求、边缘情况和更高风险场景对其进行测试。模拟和评分器检查它是否达到了正确的结果、遵循了政策、正确使用了工具，并在适当时进行了上报。当交互超出公司边界时，护栏可以进行干预。
这项工作不会在上线后停止。生产会话、上报和质量信号可以向团队展示智能体在哪些方面表现良好，以及在政策、产品和用户行为变化时哪些方面需要关注。使用Presence插件的Codex会调查这些信号并提出更新建议。团队可以针对生产中的版本测试每个提议的更改，然后批准受控的发布。
Presence还设计为与你共同学习——随着你的业务、客户和员工的发展而改进，同时企业保持控制权。它建立在OpenAI世界级的研究基础上，随着我们的模型和能力提升而不断进步。当用例超出产品当前支持的范围时，OpenAI的前沿部署工程师（FDE）和合作伙伴可以与客户合作，将其投入生产。
可用性
OpenAI Presence通过有限通用可用性计划，作为已部署产品提供给符合条件的客户。部署由OpenAI前沿部署工程师和选定的全球系统集成商主导。Presence目前尚不能作为自助服务产品使用。
在我们推出OpenAI Presence的同时，我们将继续通过OpenAI API为语音客户提供对我们前沿模型的访问。
要探索Presence是否适合您的组织，请联系您的OpenAI客户团队。
The challenge for enterprises is no longer proving that AI agents can work, it’s making them reliable enough to do high-value work in production. Agent behavior must also adapt as products, policies, and user behavior change. That requires more than a model: it requires the systems, evaluations, and deployment expertise to improve agents as those conditions change without giving up control.
We’re introducing OpenAI Presence, a battle-tested product that helps enterprises deploy trusted AI agents that can answer questions, resolve issues, use company systems, take approved actions, and escalate to people when needed. Proven through years of working with customers at enterprise-scale, Presence pairs model reasoning with policies, guardrails, and escalation rules that verify accuracy and performance.
Each deployment starts with a specific job, such as resolving billing issues, supporting insurance claims, or resolving employee IT service requests. The agent receives only the knowledge and system access required for that job. The company sets the policies: what the agent can do, when it needs approval, and when a person should take over. After launch, production sessions and escalations reveal gaps. Codex proposes updates that teams can test and approve, helping the agent adapt as customer behavior changes.
OpenAI works alongside each customer to identify each high-value workflow, connect the necessary knowledge and systems, establish permissions and policies, test the agent, and bring it into production. As the deployment expands, OpenAI and select systems integrators can continue to support it. Presence was developed in tight collaboration with OpenAI’s Research team, with generalized insights from every deployment informing ongoing research and product development and improving the product for all customers over time.
Available today for voice and chat agents
Today, Presence supports real-time experiences across voice and chat, such as customer support, outbound sales, and high-risk internal workflows. A customer might use it to resolve a billing issue—from understanding the request and verifying the customer to looking up account information, applying company policy, and taking an approved action.
Companies decide what remains consistent across deployments—such as policies, evaluations, and escalation rules—and what should change for each workflow or channel. This lets teams build on what works and expand to new use cases without starting over.
Presence brings together the components teams need to run agents in production: policies and standard operating procedures, guardrails, approved actions, simulations, evaluation tools, and a Codex-powered improvement process.
Together, these components help teams connect company systems, define how agents should behave, evaluate performance, enforce policies, and manage changes after launch.
Proven in leading enterprises
Built through years of deploying agents with customers, Presence is an ambitious product shaped by the demands of mission-critical environments. Every deployment generated insights that continuously inform OpenAI research and product development, creating compounding benefits for customers.
Presence powers OpenAI’s English-language phone support channel at 1-888-GPT‑0090, handling open-ended requests, verifying callers, using account context, and taking approved actions. Within weeks, it met or exceeded benchmarks we use to grade frontline human-support quality and now resolves 75% of inbound issues without human assistance. Working with our launch team, its Codex-powered improvement loop reduced human handoffs by 15 percentage points in just 10 days.
Leading enterprises are also building on the same proven foundation:
- BBVA is exploring AI-powered voice support for everyday banking needs in Mexico.
- SoftBank is testing natural Japanese-language customer conversations.
- IAG is exploring timely support during high-demand events such as severe weather.
1 of 3
Built for trust before, during, and after launch
Before a Presence deployment reaches users, teams can test it against common requests, edge cases, and higher-risk scenarios. Simulations and graders check whether it reached the right outcome, followed policy, used tools correctly, and escalated when appropriate. Guardrails can intervene when an interaction moves outside the company’s boundaries.
That work does not stop at launch. Production sessions, escalations, and quality signals can show teams where an agent is working well and where it needs attention as policies, products, and user behavior change. Codex using the Presence plugin investigates those signals and suggests updates. Teams can test each proposed change against the version in production, then approve a controlled rollout.
Presence is also designed to learn with you—improving as your business, customers, and employees evolve while the business stays in control. Built on OpenAI’s world-class research, it continues to advance as our models and capabilities improve. When a use case goes beyond what the product supports today, OpenAI Forward Deployed Engineers (FDEs) and partners can work with the customer to bring it into production.
Availability
OpenAI Presence is available to eligible enterprise customers as a deployed product through a limited general availability program. Deployments are led by OpenAI Forward Deployed Engineers and select global systems integrators. Presence is not yet available as a self-serve product.
As we introduce OpenAI Presence, we'll continue supporting voice customers with access to our frontier models through the OpenAI API.
To explore whether Presence is right for your organization, contact your OpenAI account team.
本文内容采集自官方网站，排版和翻译可能与原页面存在差异。
阅读官方全文