Applied AI Agent Expert / Architect

该职位来源于猎聘 About DBS Bank DBS is a leading financial services group in Asia, committed to making banking joyful and impactful for customers. We are transforming the future of banking through innovation and technology, empowering individuals and businesses to thrive in the digital economy. Job Summary We are seeking a Senior AI Agent Expert / Architect to lead the design and implementation of next-generation Multi-Agent systems for high-impact financial workflows, such as investment research, risk management, and automated auditing. In this role, you will leverage top-tier overseas commercial LLMs—with a heavy emphasis on the Google Gemini ecosystem—to build highly reliable, autonomous, and deterministic AI agents capable of solving complex, multi-step financial problems. Responsibilities

• Agent Architecture Design: Lead the top-level architecture design of production-grade Multi-Agent systems, ensuring high scalability, low latency, and robust system stability.
• Complex Workflow Orchestration: Deconstruct intricate financial business logic into autonomous agent workflows equipped with advanced planning, ReAct, and Plan-and-Solve capabilities.
• Google AI Ecosystem Optimization: Deeply integrate and fine-tune Google Gemini models via their SDKs, maximizing the utilization of native long-context windows, multimodal inputs, and context caching features.
• Toolchain & API Integration: Seamlessly connect AI agents with enterprise financial databases, quantitative tools, and real-time market data APIs, optimizing Tool/Function Calling success rates.
• Determinism & Guardrails: Implement financial-grade safety guardrails to ensure agents behave predictably, safely, and in full compliance with rigorous financial regulations. Requirements
• Core Experience: 5+ years of experience in AI/NLP or backend architecture, with at least 2 years of proven track record deploying complex LLM agents or workflow systems in a production environment.
• Agent Framework Proficiency: Deep expertise in mainstream agent orchestration and graph frameworks such as LangGraph, CrewAI, or AutoGen, with the ability to customize core components.
• Google Ecosystem Expertise: o Highly proficient with Google Vertex AI SDK or Google GenAI SDK. o Hands-on engineering experience using Gemini's native features like Context Caching and advanced Function Calling to optimize token costs and inference latency.
• Engineering Excellence: Production-level mastery of Python, with strong skills in asynchronous programming (asyncio) and high-concurrency distributed system design.
• Supporting Tech: Strong familiarity with advanced RAG (Retrieval-Augmented Generation) and vector databases to feed agents precise, domains-specific knowledge.
• Education: Bachelor’s degree or higher (Master's preferred) in Computer Science, Data Science, Mathematics, Financial Engineering, or a related quantitative field. Preferred Qualifications
• Prior professional experience in FinTech, top-tier brokerages, investment banks, or quantitative hedge funds.
• Hands-on experience with Google Vertex AI Agent Builder or enterprise safety components (e.g., Grounding with Google Search).
• Experience with LLM Fine-tuning to improve model accuracy for specific financial tools, domain vocabularies, or structured API outputs.
• Strong financial domain knowledge or professional credentials (e.g., CFA, FRM).