内容来源：https://news.microsoft.com/source/latam/features/ai/costa-rica-dos-pinos-ai-agents-microsoft-copilot-en/?lang=en

内容总结：

中美洲乳业巨头Dos Pinos：部署80余个AI代理，实现包装审核零误差与效率飞跃

在哥斯达黎加，中美洲最大乳制品合作社之一Dos Pinos正通过大规模部署人工智能代理（AI Agent）重塑其日常运营。过去，设计团队在收到新的包装设计方案时，最头疼的环节莫过于逐行核对密密麻麻的营养成分表，检查数字是否匹配、逗号是否遗漏。如今，设计师只需在自建的AI工具中输入指令，数秒内即可完成技术文档与最终标签的比对，并精准标记出所有差异。

这一名为“AI质检员”的工具由该合作社创意主管霍汉·罗德里格斯基于微软Copilot Studio自行搭建。自2025年底上线以来，包装设计阶段的信息不一致率已降至接近零，不仅优化了工作流程，还显著降低了因信息错误可能引发的监管处罚风险。罗德里格斯表示：“错误从来不在图形设计，而在于营养和法规信息——这些细节必须精确到毫厘之间。”

Dos Pinos的首席信息官亚历杭德罗·阿格达斯指出，AI不只是工具，更是合作社与员工的竞争力核心。他强调：“AI帮助我们转变工作方式，降低成本，提高效率，实现过去难以完成的任务。”目前该合作社已部署约80个AI代理，覆盖包装审核、法律文件起草、风险评估、客户互动等多个职能领域，并启动了“AI大使”计划，鼓励和赋能各岗位员工自主构建代理工具。

以销售分析师兰德尔·贝纳维德斯为例，他过去每天需花费四小时手动处理数据、分发报表。通过学习使用微软Copilot与Power BI服务，他实现了报告自动分发，将时间解放出来专注于数据分析和战略指标构建。他说：“‘分析师’的岗位不该是重复劳动。”

包装AI代理的应用效果尤为显著：设计阶段的不一致率下降超过50%，产品平均上市时间缩短约10天。信息技术与数字化转型主管卡洛斯·桑迪表示：“在制造业，10天意味着巨大的运营敏捷性提升。”

目前，Dos Pinos已在初期阶段为超过1000名员工部署了Copilot，并计划在2026年推广至全组织约6000名员工。AI负责人阿赫尼斯·马塔里塔表示：“经验告诉我们，这并不复杂。只要指令清晰、信息准确，就能快速构建出有价值的工具。”团队正进一步开发优化配送流程和制造数据信息化的新代理。

对于一线员工而言，AI带来的不仅是效率提升，更是工作心态的转变。罗德里格斯感慨：“团队曾因难以捉摸的错误而焦虑不安，现在我们对创意工作的信心更强了。工作流程更顺畅，也让我非常安心。”

中文翻译：

当新的包装改版任务交到团队手中时，霍汉·罗德里格斯（Jhojan Rodríguez）不再为工作中最棘手的部分而紧张不安。
他的设计师团队如今不再逐行扫描密密麻麻的营养成分表——去寻找一个对不上的数字或一处缺失的逗号——而是通过罗德里格斯自己构建的AI智能体输入一段提示指令。几秒钟后，它就能标记出技术文档与最终标签之间的任何差异。
“对我们来说，错误从来不在图形设计方面，”罗德里格斯说，他是双峰乳业生产商合作社（Cooperativa de Productores de Leche Dos Pinos）设计部门的创意负责人。“错误都出在营养信息和法规信息上。而这些信息必须精确到毫米级。”
这个被他称为“AI质检员”的智能体，是基于微软Copilot Studio构建的。它会在文件离开设计团队之前，将包装标签与内部技术参数表进行比对并标出差异。自2025年底投入使用以来，这些不一致之处已几乎降至零，从而优化了工作流程，并显著降低了受到监管处罚的风险。
这个包装质检员并非孤例，而是中美洲最大的乳业合作社之一——双峰公司（Dos Pinos）内部更广泛战略的体现。在整个组织中，高管和员工（双峰称之为“协作型人才”）正致力于将企业打造成前沿公司——即大规模地将AI嵌入日常工作的公司。
该合作社表示，已在各职能部门部署了约80个AI智能体，涵盖包装审核、法律文书起草、风险评估和客户互动等领域。它还推出了一个“AI大使”计划，鼓励并赋能各岗位员工构建属于自己的智能体。
“我们在AI及基于智能体的技术中看到了巨大的价值和潜力，”双峰公司首席信息官亚历杭德罗·阿格达斯（Alejandro Arguedas）说。“我们不仅将AI视为一种工具，更将其看作是我们员工和合作社的一项竞争优势。它帮助我们转变工作方式，通过降低成本、提高效率以及实现过去复杂繁琐的事情，来改进不同的流程。”
双峰公司总部位于哥斯达黎加的阿拉胡埃拉（Alajuela），拥有约6000名员工，业务涵盖乳制品生产、加工、包装、农业工业服务、物流和零售分销。牛奶日产量约130万升，来自大约1500个成员牧场，其中许多是中小型生产者，供应着种类丰富的乳制品以及农牧投入品。
在如此规模下，身处利润率低、竞争激烈且监管审查持续不断的行业，即便是微小的错误也可能波及整个运营。例如，一个包装错误就可能延迟生产并推迟产品上市；错误的营养声明可能引发法律和声誉风险，包括潜在的召回。
阿格达斯表示，对双峰公司而言，采纳AI是一项保持竞争力并确保业务长期存续的战略。“我们身处大众消费市场，毫无疑问在成本、自动化、服务、流程转型和产品质量方面面临着巨大压力。我们希望在保持领导地位的同时变得更强。”

智能体生态系统

过去一年里，双峰公司一直专注于构建一个不断壮大的、功能范围精准的AI智能体生态系统。
这些智能体使用微软AI工具Microsoft 365 Copilot、Copilot Chat和Copilot Studio开发，旨在模拟合作社内部特定的角色和任务——增强员工的日常工作能力。
除了包装质检员之外，其他智能体还支持保密协议生成、IT服务请求处理、风险文档管理，甚至还有一个“虚拟兽医”，用于向各零售点的供应商推荐畜牧产品。
在某些情况下，员工将Copilot Chat作为学习工具，以实现内部现有流程的自动化。
例如，销售分析师兰德尔·贝纳维德斯（Randall Benavides）将Copilot当作教练，一步步学习如何通过Power BI服务（Power BI Service）经由Microsoft Outlook发送销售报告。Power BI服务是微软的云平台，员工可在该平台上发布、共享并与整个组织的报告和数据表格进行交互。
“我过去每天花四个小时仅仅在处理数据上，”贝纳维德斯回忆道，他描述了这种占用了大部分时间、只给更有价值的分析任务留下极少空间的重复性工作。“当你听到‘分析师’这个词时，可不会想到是干这个的。”
IT战略与数字化转型主管卡洛斯·桑迪（Carlos Sandí）表示，双峰公司的核心目标之一是通过为员工配备能扩展其能力的AI工具，让他们能够专注于更高价值、更具战略性的工作，从而提高生产力。
“我们推出AI不是为了赶时髦，”他补充道。“AI是一种转型工具，但真正的驱动力是赋能‘协作型人才’去创新。”
该合作社表示，成果已经显现。
双峰公司称，自包装智能体投入使用以来，设计阶段的不一致之处减少了50%以上，帮助将平均上市时间缩短了约10天。“在生产方面，10天是相当长的时间，”桑迪说。“将这个时间从周期中砍掉，使我们的运营敏捷性大大提高。”
双峰公司已在第一阶段向超过1000名员工推广了Copilot，并计划在2026年全面推广至整个组织。
“我们学到的是，这不必很复杂，”人工智能主管阿赫尼斯·马塔里塔（Argenis Matarrita）说。“有了好的提示词和正确的信息，你很快就能构建出有用的东西。他的团队目前正在开发新的智能体，以优化配送流程并将来自制造设施的信息系统化。”

在整个组织推广采用

为推动并维持应用，双峰公司创建了一个内部“AI大使”计划：一个由15名员工（人数仍在增加）组成的团队，他们在各自部门担任倡导者，并帮助同事进行AI基础知识的培训。
贝纳维德斯就是其中之一。
当他最初加入销售团队时，他的日常工作被重复性手工劳动所主导：下载营收数据、筛选电子表格、一封封地给客户发邮件发送Microsoft Excel报告。每天向超过100名内部用户提供多达21份日报，每天上午就要花掉大约四个小时。
在参加了一次内部AI研讨会后，他向Copilot求助，寻求通过Power BI服务自动分发报告的指导。“我对AI一无所知。我告诉Copilot，‘我不了解这个工具——请一步步引导我。’结果成功了，”贝纳维德斯说。“工作量大大减轻了。”
得益于从数据处理中解放出来的时间，他现在专注于数据分析和构建指标，以及开发新的Power BI仪表板。他还支持其他团队扩展Copilot的应用。
罗德里格斯在构建包装智能体时也有类似的经历。
使用Copilot Studio，他用大约250个词定义了任务，并给出了清晰的指令，说明智能体应如何评估信息。该工具不会自动进行更改，而是为设计师提供详细的分析，以便他们据此采取行动。
“这并非要取代我们的设计师，”他说。“而是要移除工作中最机械、最容易出错的部分，这样她才能专注于设计质量。”
在最近的一个项目中，罗德里格斯团队的一名设计师使用该智能体审核了一个更新的奶粉包装，同时上传了技术规格书和拟议的包装设计稿。智能体生成了详细的对比报告，识别出差异并根据内部参数（从错误数字到格式不一致之处）推荐了修正方案。
对罗德里格斯来说，最具意义的影响已超越了效率。过去，团队曾因难以捉摸的错误而感到压力和担忧，如今他们工作起来信心大增。
“工作流程更顺畅了，”他说。“这让我安心了很多。现在我们能充满信心地专注于创意方面了。”

胡安·蒙特斯（Juan Montes）撰文探讨AI和数字创新如何重塑拉丁美洲和加拿大的行业格局与决策方式。他的报道涵盖从跨国公司为高管部署AI智能体，到公立学校教师在课堂上应用技术等故事。他出生于马德里，曾在西班牙和危地马拉担任记者，并担任过《华尔街日报》驻墨西哥、中美洲和加勒比地区的海外特派记者。您可以在领英（LinkedIn）上联系他。
主图：双峰公司拥有约6000名员工，业务涵盖乳制品生产、加工、包装和农业工业服务。图片来源：双峰公司。

英文来源：

When a new packaging redesign lands on his team, Jhojan Rodríguez no longer braces for the most stressful part of the job.
Instead of scanning dense nutritional tables line by line—looking for a mismatched number or a missing comma—the designer on his team now runs a prompt through an AI agent Rodríguez built himself. Seconds later, it flags any discrepancies between the technical documentation and the final label.
“For us, the errors were never about the graphic aspect,” says Rodríguez, creative lead of Cooperativa de Productores de Leche Dos Pinos’ design department. “They were in the nutritional and regulatory information. And those must be precise down to the millimeter.”
This “AI inspector,” as Rodríguez calls it, was built using Microsoft Copilot Studio. The agent compares packaging labels against internal technical sheets and highlights discrepancies before files leave the design team. Since its launch in late 2025, these inconsistencies have been reduced to nearly zero, improving workflows and significantly lowering the risk of regulatory penalties.
The packaging inspector isn’t a one-off, but rather evidence of a broader strategy within Dos Pinos, one of Central America’s largest dairy cooperatives. Across the organization, executives and employees, whom Dos Pinos calls “collaborative people,” are working to evolve into a frontier company—one that embeds AI into everyday workflows at scale.
The cooperative says it has already deployed around 80 AI agents across functions, from packaging review and legal drafting to risk assessments and client engagement. It has also launched an “AI ambassadors” program to encourage and equip employees across roles to build their own agents.
“We see enormous value and potential in AI and agent-based technologies,” says Alejandro Arguedas, Dos Pinos’ chief information officer. “More than just a tool, we see AI as a competitive advantage for both our people and the cooperative. It helps us transform how we work and improves different processes by reducing costs, increasing efficiency and enabling things that were previously complex.”
Based in Alajuela, Costa Rica, Dos Pinos has about 6,000 employees and operations spanning dairy production, processing, packaging, agro-industrial services, logistics and retail distribution. Milk production—some 1.3 million liters per day—comes from roughly 1,500 member farms, many of them small and medium-sized producers, supplying a wide portfolio of dairy products as well as agricultural and livestock inputs.
At that scale, in a business where margins are tight, competition is strong and regulatory scrutiny is constant, even small errors can ripple across operations. A packaging mistake, for instance, can delay manufacturing and push back product launches; incorrect nutritional declarations can trigger legal and reputational risks, including potential recalls.
For Dos Pinos, adopting AI is a strategy to stay competitive and secure long-term business viability, Arguedas says. “We operate in a mass consumer market, and there is no doubt there is enormous pressure around costs, automation, service, process transformation and product quality. We want to maintain our leadership and become even stronger.”
An ecosystem of agents
Over the past year, Dos Pinos has focused on building a growing ecosystem of narrowly scoped AI agents.
Developed using the AI tools Microsoft 365 Copilot, Copilot Chat and Copilot Studio, the agents are designed to mirror specific roles and tasks inside the cooperative—augmenting employees in their day-to-day work.
Alongside the packaging inspector, other agents support the creation of non-disclosure agreements, IT service request handling, risk documentation management and even a “virtual veterinarian” that recommends livestock products to vendors across retail locations.
In some cases, employees use Copilot Chat as a learning tool to automate existing internal processes.
Sales analyst Randall Benavides, for example, used Copilot as a coach to learn step by step how to send sales reports through Microsoft Outlook via Power BI Service, Microsoft’s cloud platform where staff publish, share and interact with reports and data tables across the organization.
“I was spending four hours a day just processing data,” Benavides recalls, describing repetitive work that took up most of his time and left little room for more valuable tasks like analysis. “That’s not what you expect when you hear the word ‘analyst.’”
A core goal for Dos Pinos is to increase productivity by equipping staff with AI tools that expand their capabilities and allow them to focus on higher-value and more strategic work, says Carlos Sandí, head of IT strategy and digital transformation.
“We’re not rolling out AI just because it’s trendy,” he adds. “AI is a tool for transformation, but the real driver is empowering the collaborative people to innovate.”
The results are already visible, the cooperative says.
Since the packaging agent was implemented, inconsistencies in the design stage have decreased by more than 50%, Dos Pinos says, helping reduce the average time to market by roughly 10 days. “In production terms, 10 days is a huge amount of time,” Sandí says. “Cutting that from the cycle makes us far more operationally agile.”
Dos Pinos has rolled out Copilot to more than 1,000 employees in an initial phase, with plans to expand across the entire organization throughout 2026.
“What we’ve learned is that this doesn’t have to be complex,” says Argenis Matarrita, head of AI. “With good prompts and the right information, you can build something useful very quickly.” His team is now working on new agents to optimize delivery processes and systematize information from manufacturing facilities.
Scaling adoption across the organization
To drive and sustain adoption, Dos Pinos has created an internal “AI ambassadors” program: a group of 15 employees and counting, who act as champions within their departments and help train colleagues on AI fundamentals.
Benavides is one of them.
When he first joined the sales team, his days were dominated by repetitive manual work: downloading revenue data, filtering spreadsheets and emailing Microsoft Excel reports to clients one by one. Delivering up to 21 daily reports to more than 100 internal users took about four hours each morning.
After attending an internal AI workshop, he turned to Copilot for guidance on automating report distribution through Power BI Service. “I knew nothing about AI. I told Copilot, ‘I don’t know this tool—guide me step by step.’ And it worked,” Benavides says. “The workload eased up a lot.”
He now focuses on data analysis and building indicators, as well as developing new Power BI dashboards, thanks to the time freed up from data processing. He also supports other teams in expanding Copilot adoption.
Rodríguez had a similar experience when building the packaging agent.
Using Copilot Studio, he defined the task in about 250 words, with clear instructions for how the agent should evaluate information. The tool does not make changes automatically; instead, it provides a detailed analysis for designers to act on.
“This isn’t about replacing our designer,” he says. “It’s about removing the most mechanical, error-prone part of the job so she can focus on design quality.”
On a recent project, a designer on Rodríguez’s team used the agent to review an updated powdered milk package, uploading both the technical specifications and the proposed artwork. The agent produced a detailed comparison, identifying discrepancies and recommending corrections based on internal parameters—from incorrect figures to formatting inconsistencies.
For Rodríguez, the most meaningful impact goes beyond efficiency. Where his team once felt stressed and concerned about elusive errors, they now work with greater confidence.
“Work flows better,” he says. “It’s given me a lot of peace of mind. Now we can focus on the creative side with confidence.”
Juan Montes writes about how AI and digital innovation are reshaping industries and decision making across Latin America and Canada. His reporting spans stories from multinational companies deploying AI agents for executives to public school teachers adopting technology in classrooms. Born in Madrid, he worked as a journalist in Spain and Guatemala and was a foreign correspondent for the Wall Street Journal in Mexico, Central America and the Caribbean. You can contact him on LinkedIn.
Top image: Dos Pinos has about 6,000 employees and operations spanning dairy goods production, processing, packaging and agroindustrial services. Photo by Dos Pinos.