AI Agent 高级研发工程师 / 落地实施专家，FTO

Information Technology • Full Time • Hong Kong, Hong Kong, HK • 13h ago

创造非凡影响的职业生涯

秉承“以人为本”的发展理念，我们致力于成为培养未来金融科技领军人才的摇篮。在这里，我们与时代并进，为您提供具有行业竞争力的研发平台与多元化的职业发展路径。我们鼓励团队成员在利用前沿技术创造真实商业价值的过程中，获得成就感与幸福感。我们致力于打造一个包容、开放且充满活力的科技与金融融合环境，赋予人才更多的自主权，助您在实现前沿技术落地的同时，达成理想的工作与生活平衡。

广阔的机遇世界正向您敞开。加入我们，开启您的 AI 探索之旅。

关于我们的业务

在快速演进的数字化浪潮中，我行金融科技团队的视野早已超越传统的银行业务边界。作为前沿技术与真实业务之间的“AI 使者”，我们就像是一个现代化的信息与算力驿站，致力于将生成式人工智能（Gen-AI）与大语言模型（LLM）从“概念对话”转化为深度的“生产力引擎”，并精准地传递、赋能给内部员工与终端客户。

作为 AI Agent 实施专家，您将成为推动 Gen-AI 技术在银行真实业务场景中落地的核心力量。您不仅将提升业务运转的效率，更将与业务侧的行业专家紧密协作，攻克复杂的工程难题，不断拓宽数字金融的边界。

您的核心工作内容

● 分析与评估：深入理解银行业务痛点，评估 AI Agent 应用场景，并主导以用户体验和业务结果为导向的 AI 解决方案落地。

● 架构与开发：设计并实施具备感知、规划（Planning）、记忆及工具调用（Tool Use）能力的 AI Agent 核心闭环。根据业务复杂度，灵活运用高代码框架（如 LangGraph 等）构建严谨的复杂业务流，或采用低代码平台（如 Dify、Coze）实现敏捷交付与快速验证。建立工程化AI应用开发的过程管理规范，保证大规模AI应用开发上线的质量和运维可靠性。

● 知识检索赋能：主导企业级 RAG（检索增强生成）系统的建设，精通至少一种主流向量数据库（如 Milvus, Qdrant 等）的应用与性能调优，从根本上解决金融场景下的长文本解析与模型幻觉问题。

● 系统融合：作为技术桥梁，将创新的 AI Agent 逻辑与银行既有核心系统（Legacy Systems）及 API 进行安全、稳定的集成。

● 业务协同：在整个项目生命周期中，担任业务与技术团队的联络枢纽，用清晰的逻辑向各级管理层汇报 AI 技术路径与实操进展，确保项目的成功交付。

在加入我们的这段时间里，您将在以下领域发挥才能并进一步发展您的专业技能与领导力：

敏捷交付驱动 (Driving agility)、构建多元化技术能力 (Building diverse capability)、跨部门协作 (Collaborating)、持续交付价值 (Delivering value)、复杂系统的质量与风险管理 (Managing quality & risk)、以及卓越的技术敏锐度 (Tech savviness)。

我们期待这样的您

● 计算机科学、软件工程、数学或相关理工科专业的本科及以上学历。

● 具备至少 1 年以上 的 LLM/Agent 实际应用开发经验。

● 卓越的动手实操与工程能力：相较于学术论文的发表，我们高度关注您的 GitHub 贡献记录或个人开源项目库。面试中我们将期待您展示亲自编写的 AI Agent，或探讨您在解决实操难题时的独到见解。

● 精通 Python 或 Java，深入理解 Function/Tool Calling 机制，并具备处理结构化与非结构化数据的扎实功底。

● 对系统稳定性、数据安全及金融行业的合规要求有一定了解与实践经验。

● 具备强烈的技术好奇心、自我驱动力以及在快节奏、多优先级环境下的适应能力。

● (加分项) 拥有在开源社区（如 GitHub 500+ Stars、Hugging Face）的活跃贡献记录，或具备大模型私有化本地部署（如 vLLM, Ollama）经验者优先。

Your Core Responsibilities

● Analysis and Evaluation: Deeply understand banking business pain points, evaluate AI Agent application scenarios, and lead the implementation of AI solutions oriented towards user experience and business outcomes.

● Architecture and Development: Design and implement the core AI Agent closed-loop system with perception, planning (Planning), memory, and tool use (Tool Use) capabilities. Flexibly utilize high-code frameworks (e.g., LangGraph) to construct rigorous complex business flows based on business complexity, or adopt low-code platforms (e.g., Dify, Coze) to achieve agile delivery and rapid validation. Establish process management standards for engineering AI application development, ensuring the quality and operational reliability of large-scale AI application deployment.

● Knowledge Retrieval Empowerment: Lead the construction of enterprise-level RAG (Retrieval-Augmented Generation) systems, master the application and performance optimization of at least one mainstream vector database (e.g., Milvus, Qdrant), and fundamentally resolve long-text parsing and model hallucination issues in financial scenarios.

● System Integration: Act as a technical bridge to securely and stably integrate innovative AI Agent logic with the bank’s existing core systems (Legacy Systems) and APIs.

● Business Collaboration: Serve as the central liaison between business and technical teams throughout the project lifecycle, clearly communicating AI technical approaches and operational progress to management at all levels, ensuring project success delivery.

We Expect You to Have:

● A bachelor’s degree or higher in Computer Science, Software Engineering, Mathematics, or related technical fields.

● At least 1 year of hands-on experience in LLM/Agent application development.

● Exceptional practical and engineering capabilities: We focus more on your GitHub contribution records or personal open-source projects than academic paper publications. During interviews, we expect you to demonstrate AI Agents you have personally developed or discuss your unique insights into solving practical challenges.

● Proficiency in Python or Java, in-depth understanding of Function/Tool Calling mechanisms, and solid expertise in handling structured and unstructured data.

● Understanding and practical experience in system stability, data security, and compliance requirements in the financial industry.

● Strong technical curiosity, self-motivation, and adaptability in fast-paced, multi-priority environments.

● (Preferred) Active contribution records in open-source communities (e.g., GitHub 500+ Stars, Hugging Face), or experience in private on-premises deployment of large models (e.g., vLLM, Ollama).

Similar Jobs

Apply