#132811 - 软件工程师
#132811 - Software Engineer
我们正在寻找一位动手能力强的软件工程师，负责构建和运营帮助AI代理访问并基于企业系统上下文进行推理的产品服务。这是一项日常软件开发工作，核心是Model Context Protocol（MCP）服务和生产产品工程。该职位不涉及ETL管道、系统配置或管理，也不涉及独立脚本开发。
主要职责
• 构建和运营基于MCP的服务，扩大AI代理可访问的上下文。
• 提供可靠方案用于托管和管理MCP服务。
• 构建一个只读MCP服务，从企业协作、生产力和开发者系统中暴露已批准的上下文。
• 在多个企业系统和信息源之间集成AI代理的上下文访问。
• 仔细阅读并理解生产代码。
• 在代码审查期间向团队成员提供实质性反馈。
• 直接负责从开发到生产的所有实现、故障排查和交付工作。
必备技能
• 具备5年以上中级水平的软件开发经验。
• 具有实际的AI和代理开发经验。
• 对AI原理及其底层技术有深刻理解。
• 具有Java或Go的实际开发经验，优先考虑Java。
• 具有在软件工程工作流中使用Git的经验。
• 强大的问题解决能力和强烈的技术责任感。
• 有端到端开发软件产品的成功经验。
• 当前具备编写、阅读和理解生产代码的能力。
• 有企业经验者优先。
加分项技能
• 熟悉Model Context Protocol的设置和配置。
• 了解常见的MCP故障排除场景。
• 有负责和维护MCP服务器的经验。
• 有构建获得广泛采用的代理技能，并加速用户或团队的经验。
• 具备系统思维能力，能够将问题分解为清晰、可组合的抽象层次。
所需工具与平台
• Java或Go。
• Git。
• AI和代理开发技术。
• 熟悉Model Context Protocol者优先。
地点、时间与参与方式
• 候选人必须位于拉丁美洲（LATAM）。
• 需要完全覆盖美国中部时间的工作时间。
• 这是一个每周40小时的全职合同职位。
• 预计参与将持续至2027年3月31日。
• 目前这不是一个合同转正的机会。
查看英文原文
We are seeking a hands-on Software Engineer to build and operate product services that help AI agents access and reason over context from enterprise systems. This is a day-to-day software development role centered on Model Context Protocol services and production product engineering. It is not focused on ETL pipelines, systems configuration or management, or standalone script development.
Key Responsibilities
• Build and operate MCP-based services that broaden the context available to AI agents.
• Deliver a reliable solution for hosting and managing MCP services.
• Build a read-only MCP service that exposes approved context from enterprise collaboration, productivity, and developer systems.
• Integrate AI-agent context access across multiple enterprise systems and information sources.
• Read and understand production code thoroughly.
• Provide substantive feedback to teammates during code reviews.
• Take direct ownership of implementation, troubleshooting, and delivery from development through production.
Must-Have Skills
• 5+ years of software development experience at the mid-level.
• Hands-on AI and agentic development experience.
• Strong understanding of AI principles and their underlying technologies.
• Hands-on development experience in Java or Go, with Java preferred.
• Experience using Git in a software engineering workflow.
• Strong problem-solving skills and a strong sense of technical ownership.
• Proven experience developing a software product end to end.
• Current hands-on ability to write, read, and understand production code.
• Enterprise experience strongly preferred.
Nice-to-Have Skills
• Familiarity with Model Context Protocol setup and configuration.
• Understanding of common MCP troubleshooting scenarios.
• Experience owning and maintaining an MCP server.
• Experience building agent skills that achieved broad adoption and accelerated users or teams.
• Strong systems thinking and the ability to break problems into clear, composable abstraction layers.
Required Tools & Platforms
• Java or Go.
• Git.
• AI and agentic development technologies.
• Model Context Protocol familiarity is preferred.
Location, Time & Engagement
• Candidates must be based in LATAM.
• Full working coverage of US Central Time is required.
• This is a full-time contract engagement at 40 hours per week.
• The anticipated engagement runs through March 31, 2027.
• This is not currently a contract-to-hire opportunity.