Building OpenAI with OpenAI

首席商务官 Giancarlo “GC” Lionetti 拉开了我们关于 OpenAI 如何在自身业务中基于我们技术构建解决方案的系列序幕。

人工智能已不再是试验性事物。它如今成为工作的基础设施，从试点项目走向影响日常决策的系统。尽管我们的模型在速度、成本和能力上不断提升，但采用往往并非一路顺畅。部署常常跑在组织为充分利用该技术所需变革之前。

在 OpenAI 内部我们也感受到同样的紧张。把业务建在 AI 之上，意味着要面对每个客户都会问的问题：从何处开始，如何将新工具与现有工作流对齐，如何在环境不断变化时衡量进展。当我与客户会面时，他们都会问：“OpenAI 自己如何使用 OpenAI？”

我们的做法是把 AI 当作一种提升职能的实践。

每家公司都依赖专业技能。建立信任的销售人员、解决最难问题的支持负责人、在复杂中寻找秩序的工程师。AI 将这些专业知识编码并分发到各个团队，放大每种职能的影响力。

这就是我们的构建方式。我们的 GTM（市场拓展）、产品和工程团队研究日常工作流，定义优秀的标准，并在数周而非数季度内交付变更。我们决定专注于若干具有高杠杆且影响巨大的系统。每个团队都在真实部署中测试这些系统，锻炼与我们的客户相同的能力。

OpenAI 在 OpenAI

今天我们发布 “OpenAI 在 OpenAI” 系列，展示我们如何在自身业务中使用 AI。每个故事涵盖一个真实问题以及我们构建的解决方案。我们的目标是分享可供企业借鉴的模式。

我们从几个示例开始：

• GTM Assistant（GTM 助手，基于 Slack 的工具）：将客户帐户上下文和专家知识集中在一起。它简化了研究、会议准备和产品问答，提高销售生产力并改善结果。（https://openai.com/index/openai-gtm-assistant/）
• DocuGPT：一个把合同转化为结构化、可搜索数据的智能代理。财务团队用它实现更快、更一致的大规模审查。（https://openai.com/index/openai-contract-data-agent/）
• Research Assistant（研究助理）：一个将数百万条支持工单转为会话式洞见的系统。团队能够在分钟而非数周内发现趋势并对客户反馈采取行动。（https://openai.com/index/openai-research-assistant/）
• Support Agent（支持代理）：基于 AI 代理、持续评估和动态知识循环的运营模式。它把每次交互都变成训练数据，提高质量，并使客服人员成为系统构建者而非单纯的工单处理者。（https://openai.com/index/openai-support-model/）
• Inbound Sales Assistant（入站销售助理）：一个为每个潜在客户个性化响应、即时回答产品与合规问题，并将合格线索以完整上下文转给销售代表的系统。它把错失的机会转化为收入。（https://openai.com/index/openai-inbound-sales-assistant/）

未来工作方式的预览

每家公司都有其专业技艺。AI 可以放大这种技艺。未来属于那些能将员工的专业知识捕捉并在公司内部分发的组织。将技艺与代码结合的公司将设定前沿。

如果您想了解更多，我们很乐意与您交流。加入我们 10 月 6 日的 DevDay，随后会有技术资源发布。

准备好在您的业务中让 ChatGPT 发挥作用了吗？

联系我们的团队（https://openai.com/contact-sales/）

Chief Commercial Officer, Giancarlo “GC” Lionetti, kicks off our series about how OpenAI is building its own solutions on our technology.

AI has moved beyond an experiment. It now operates as infrastructure for work, shifting from pilots to systems that shape daily decisions. While our models improve in speed, cost, and capability, adoption rarely moves in a straight line. Deployments often outpace the change needed for organizations to leverage this technology.

Inside OpenAI we see the same tension. Running our business on AI means facing the questions every customer asks: where to start, how to align new tools with existing workflows, how to measure progress as the ground shifts. When I meet customers, the question they all ask me is, “How does OpenAI use OpenAI?”

Our approach is to treat AI as a practice that elevates craft.

Every company depends on expertise. The salesperson who builds trust, the support lead who solves the hardest problem, the engineer who finds order in complexity. AI encodes that expertise and distributes it across teams, scaling the impact of each discipline. 

This is how we build. Our GTM, product, and engineering teams study their everyday workflows, define what good looks like and deliver changes in weeks instead of quarters. We decided to focus on a few high-leverage systems with outsized impact. Each team tests them in live deployments, building the same muscles our customers do. 

OpenAI on OpenAI

Today we’re launching OpenAI on OpenAI, a series that shows how we use AI inside our business. Each story covers a real problem and the solution we built. Our goal is to share patterns companies can adapt.

We start with a few examples:

• GTM Assistant⁠: a Slack-based tool that centralizes account context and expert knowledge. It streamlines research, meeting prep, and product Q&A, boosting sales productivity and improving outcomes.
• DocuGPT⁠: an agent that converts contracts into structured, searchable data. Finance teams use it for faster, more consistent review at scale.
• Research Assistant⁠: a system that turns millions of support tickets into conversational insights. Teams surface trends and act on customer feedback in minutes, not weeks.
• Support Agent⁠: an operating model built on AI agents, continuous evals, and dynamic knowledge loops. It turns every interaction into training data, raises quality, and positions reps as system builders rather than ticket handlers.
• Inbound Sales Assistant⁠: a system that personalizes responses for every lead, answers product and compliance questions instantly, and routes qualified prospects to reps with full context. It turns missed opportunities into revenue.

A preview of the future of work

Every company has craft. AI scales it. The future belongs to organizations where employees capture their expertise and distribute it across the company.  The companies that marry craft and code will set the frontier.

If you’d like to learn more, we’d love to connect. Join us at DevDay on October 6, with technical resources to follow soon after.

Ready to put ChatGPT to work in your business?

Generated by RSStT. The copyright belongs to the original author.

Source