Tìm kiếm việc làm online mới nhất!

《關於我們》 我們是一家專注於生成式 AI 客服解決方案的技術新創團隊，致力於協助電子商務、社群平台等企業，自動化處理客服訊息、提升用戶互動體驗。我們自主開發的 AI Agent 產品已被導入多語系、跨平台應用，並與全球多家知名 SaaS 服務建立技術合作夥伴關係。 我們的總部位於台北市中心，具備成熟彈性工作制度，產品線發展穩健，正擴大工程研發團隊規模！ 《工作內容》 負責 AI Agent 核心功能開發、效能優化與架構設計 串接內外部 API 並進行系統整合 解決 LLM 呼叫與 Prompt Debug 過程中的技術挑戰 參與產品測試、自動化評估與性能監控流程 與產品團隊協作，打造穩定可擴展的 AI 產品 《技術環境》 開發語言：Python AI技術：OpenAI API、LLM Tool Call、RAG、向量資料庫 架構相關：Docker、Git Flow、PostgreSQL、SQLite 產品工具整合：Zendesk、內建客服平台 加分項目：有 AI Chatbot 開發經驗或 AWS Lambda 架構應用經驗者佳

公司介紹 我們是一支總部位於台北、放眼全球市場的量化科技團隊，專注於以自研低延遲基礎建設與演算法策略，在全球資本市場中創造穩定優勢。 團隊結合軟體工程、AI 研究與硬體加速技術（HPC／GPU／FPGA），打造奈秒級決策系統。每日處理龐大即時市場數據，並將研究成果直接轉化為可量化的交易成果。 這裡不是做 demo 或概念驗證，而是將模型真正部署在高壓、即時決策環境中。工程師能直接影響核心系統效能與策略收益，並在高度資源支持下快速實驗與迭代。 如果你期待的是： ✔ 真正落地的 AI 應用場景 ✔ 高品質數據與高算力資源 ✔ 與頂尖量化研究員並肩作戰 ✔ 成果與績效高度連動的回報機制 這將會是一個極具成長性的舞台。 工作內容 你將專注於建立與強化 LLM Agents 與 Reinforcement Learning 能力，直接應用於交易自動化與市場洞察： 與量化研究員共同設計實驗，將模型輸出轉為可交易訊號 建立 AI Agents（RAG、Tool Use、Task Planning）提升研究與交易效率 進行 Prompt Engineering、LoRA、Quantization、Distillation 等模型優化 針對金融文本（新聞、財報、社群、事件）進行 Fine-tuning 與 Alignment（SFT/DPO） 建立 Offline/Online Backtesting 與 A/B Testing 流程 導入 MLOps（CI/CD、Containerization、Model Versioning）提升模型治理能力 追蹤前沿 LLM 研究並進行內部技術分享 使用技術 Python / PyTorch / TensorFlow LLM 生態系：LangGraph、Ollama、OpenAI SDK Fine-tuning 技術：LoRA、MoE、Quantization、Distillation Alignment：SFT / DPO AI Agent Architecture（RAG、Tool Calling、Planning） HPC / GPU 加速環境 分散式數據處理與低延遲基礎建設 為什麼值得加入？ 真實高價值應用場景（模型直接影響交易決策） 高品質資料與充足算力資源 扁平透明文化，高自主權與技術決策影響力 績效高度連動的薪酬結構 國際市場布局與未來海外發展機會

HOYA BIT 致力成為加密貨幣新鮮人的最佳信任夥伴，我們致力研究加密貨幣的解決場景，並針對用戶需求提供解決方案，讓加密貨幣便利每個人的生活! ▎經營理念 享受工作與生活 當你踏入 HOYA BIT 後，你會發現我們致力讓員工創造與突破，高成就感與高報酬是讓每一個夥伴全力以赴的原動力! 高度成長與不斷創新 HOYA BIT 鼓勵每一位的夥伴不斷創新，唯有求新求變才能撼動市場，為此我們也提供每位夥伴學習經費，並透過每個月的內部交流及外部講座提升每一個夥伴的能力！ 客戶至上的服務模式 由於 HOYA BIT 是主打新手友善的加密貨幣交易所，所以我們以積極熱情的精神來服務我們的客戶，並致力透過每一個用戶的回饋來改善 HOYA BIT 的產品及服務，唯有熱情擁抱用戶，才能得到用戶的最高信任。 ▏我們相信越好的人才就需要越少的管理，所以在管理上我們只要求夥伴們尊重會議所決定的專案期程，在過程中若遇任何困難都請「即時提出」、才能「及時解決」，在辦公部份我們也致力打造讓大家都能舒適工作的環境。 ▏

HOYA BIT 致力成為加密貨幣新鮮人的最佳信任夥伴，我們致力研究加密貨幣的解決場景，並針對用戶需求提供解決方案，讓加密貨幣便利每個人的生活! ▎經營理念 享受工作與生活 當你踏入 HOYA BIT 後，你會發現我們致力讓員工創造與突破，高成就感與高報酬是讓每一個夥伴全力以赴的原動力! 高度成長與不斷創新 HOYA BIT 鼓勵每一位的夥伴不斷創新，唯有求新求變才能撼動市場，為此我們也提供每位夥伴學習經費，並透過每個月的內部交流及外部講座提升每一個夥伴的能力！ 客戶至上的服務模式 由於 HOYA BIT 是主打新手友善的加密貨幣交易所，所以我們以積極熱情的精神來服務我們的客戶，並致力透過每一個用戶的回饋來改善 HOYA BIT 的產品及服務，唯有熱情擁抱用戶，才能得到用戶的最高信任。 ▏我們相信越好的人才就需要越少的管理，所以在管理上我們只要求夥伴們尊重會議所決定的專案期程，在過程中若遇任何困難都請「即時提出」、才能「及時解決」，在辦公部份我們也致力打造讓大家都能舒適工作的環境。 ▏

AI Engineering LLM Applications | Senior AI Engineer / Tech Lead - TL (Global Financial Group | Fluent English Required) Company Overview Our client is the Taiwan entity of a well-established global financial group with operations across multiple international markets. The organization is known for its strong governance, financial stability, and long-term commitment to technology-driven transformation. In recent years, the company has made significant investments in AI and data-driven initiatives, forming a dedicated AI team focused on delivering production-ready AI solutions that directly support core business operations and customer-facing platforms. This role involves frequent collaboration with overseas teams, making strong English communication skills essential. Responsibilities Lead AI Engineers and Software Engineers to design and deliver AI-powered solutions Apply Machine Learning and LLM technologies to real-world business systems (e.g., call center, telesales, customer interaction and call routing platforms) Collaborate closely with global stakeholders to understand requirements and drive execution Design, build, and maintain scalable, reliable, and secure AI services Ensure high code quality, system performance, and long-term maintainability Contribute to technical decisions, architecture design, and best practices Balance evolving priorities while maintaining delivery momentum Technologies Python (production-grade development) Machine Learning / Deep Learning(PyTorth, TensorFlow) Large Language Models (LLMs) Generative AI FastAPI, Uvicorn Docker, Kubernetes Distributed, high-throughput, low-latency system architectures CI/CD and cloud deployment experience is a plus

公司介紹 我們的客戶為台灣具指標性的大型金融機構，長期被市場視為金融科技與 AI 落地的領先者。近年在 Generative AI 領域投入大量資源，不僅推動多項 GenAI 旗艦方案，更成功將 AI 技術導入實際營運流程，建立可複製、可擴展的企業級應用模式。 本職缺隸屬於核心的轉型單位（約百人規模），高階長官具備資工博士級背景，組織定位關鍵、決策鏈短，並直接影響 AI 產品是否能安全、穩定地交付給最終客戶。 職務說明 我們正在尋找一位 AI 模型主管，從零建立企業級 AI 品質與評測體系。 你將擔任 AI 產品上線前的 「守門員（Gatekeeper）」，全權負責 LLM、RAG 與 ML 系統的評測標準、發布流程與風險控管，平衡模型效能、安全性與合規性，對產品能否正式上線具有關鍵影響力。 工作內容 一、AI 評測策略與框架建立 從零打造企業級 AI 評測架構，定義 LLM / RAG / ML 系統的核心評估指標 評測面向涵蓋：準確率、相關性、真實性（Faithfulness）、一致性（Consistency） 將 Responsible AI 原則納入驗收標準（公平性、偏見、可解釋性） 推動評測流程自動化，並與 MLOps / CI-CD 深度整合 二、基準測試與資料治理 規劃並管理 Golden Dataset，涵蓋關鍵業務流程、Edge Cases 與對抗性樣本 追蹤並評估主流評測工具與方法（如 RAGAS、DeepEval、Arize Phoenix） 建立模型版本間可比較、可追蹤的 Benchmark 體系 三、AI 安全與紅隊演練 建立 AI 安全防護與風險驗證機制 主導 LLM 紅隊演練（Red Teaming） 制定並驗證 Prompt Injection、Jailbreaking、PII 洩露等風險的防禦與壓力測試標準 四、跨部門協作與團隊領導 與演算法、後端工程、產品團隊密切合作，建立錯誤分析回饋迴圈 針對生產環境 Bad Cases 進行 RCA，提出 Fine-tuning / Prompt Engineering 優化建議 負責團隊培育、流程標準化與評測文件制度建立

【公司介紹】這是一家隸屬大型企業集團的知名品牌，長期深耕消費者服務與企業解決方案，近年正全面推動 AI 與數位轉型。公司擁有穩定的市場基礎與龐大的用戶場景，並積極將生成式 AI 導入多元業務流程與產品服務中。目前 AI / Data 團隊由 PM、DA 與 AI Engineer 組成，採用敏捷式開發 (Scrum) 協作模式，並直接獲得高層支持。團隊不僅負責技術研究，更著重於 將 AI 真正落地到產品與商業流程，讓工程師能直接參與從架構設計到應用場景落地的完整過程，對產品與決策產生實際影響。若你期待在穩定資源環境下，挑戰真實商業場景中的 AI 應用，這會是一個能發揮技術價值與影響力的機會。 【工作內容】• 設計與開發 LLM 驅動的 AI Agent 系統架構與核心模組• 建立任務規劃、記憶管理、工具調用與流程執行等 Agent 能力• 利用 Google ADK / Vertex AI 等平台打造具備推理能力的 AI 系統• 建構並優化 RAG（Retrieval-Augmented Generation）架構，降低模型幻覺並提升知識準確度• 整合 Function Calling、Skills 等技術，建立具上下文理解與知識注入能力的智慧系統• 建立 LLM / Agent 評估機制與自動化測試工具，持續提升系統品質與可靠度• 推動 AI 在搜尋、客服、自動化分析與數據流程等場景的應用落地 【使用技術】• Python• SQL / Database Design• RESTful API• LLM Application Development• RAG Architecture• AI Agent Systems• Prompt Engineering 加分技術：• GCP（Vertex AI / Google ADK）• Vector Database（Pinecone / Weaviate / Vertex AI Vector Search）• LLM Evaluation Framework（RAGAS / DeepEval / LangSmith）

Vulcan’s mission is to keep Generative AI safe, secure, and compliant for organizations worldwide. As a Frontend Software Engineer, you’ll build intuitive, high-performance user interfaces for products like Vulcan Protect (real-time monitoring dashboards) and Vulcan Attack (security insights and controls for automated vulnerability assessments). Working with a modern frontend-focused tech stack, you’ll collaborate closely with AI security experts and backend engineers to translate complex security, compliance, and reliability challenges into clear, actionable user experiences. In this dynamic, mission-driven environment, you’ll write high-quality, maintainable frontend code, contribute to UI/UX and architectural design reviews, and help shape products that make GenAI adoption safer and smarter. We value people who are creative, speak up with their ideas, take ownership as a team player, actively engage in discussions, are eager to learn, and are easy to collaborate with. Learn more about us 👉 Vulcan product: https://vulcanlab.ai/Vulcan LinkedIn:https://www.linkedin.com/company/vulcanlab-ai/AIFT group:https://aift.io/ - How to apply Please apply with English CVthroughhttps://job-boards.greenhouse.io/aift/jobs/5744133004 Thank you! - What you'll do Focus primarily on building and maintaining features for Vulcan’s products, with an emphasis on frontend development using TypeScript, React, Next.js, React Query, Tailwind CSS, etc. Utilize additional tools and technologies as needed with GitLab CI, Docker, shell scripting, etc. Write clean, efficient, and testable code, primarily for frontend systems, with some backend contributions. Collaborate with team members during design and code reviews to build scalable, high-quality solutions, with a focus on frontend architecture. Troubleshoot and resolve technical issues by analyzing logs, system performance, and user feedback. Build and maintain internal tools to support project teams. Contribute to documentation and keep it current as our products and team grow, with an emphasis on frontend-related updates. -

We are seeking a Senior Software Engineer to be the technical heartbeat of our GenAI security capabilities. In this role, you sit at the exciting intersection of Research and Product, turning breakthrough experiments into scalable, production-grade software. You won’t just be implementing features; you’ll be designing the sophisticated tooling and AI Agent architectures that make the Vulcan Platform unique. We offer a playground for builders who want to solve unsolved problems, master advanced prompt engineering, and work alongside a world-class team. Come build the shield for the AI era and expand your technical horizons. Learn more about us 👉 Vulcan product: https://vulcanlab.ai/Vulcan LinkedIn:https://www.linkedin.com/company/vulcanlab-ai/AIFT group:https://aift.io/ - How to apply Please apply with English CVthroughhttps://job-boards.greenhouse.io/aift/jobs/5744136004 Thank you! - What you'll do 1. Core Application Platform Development (Full Stack) Vulcan Platform: Partner with Product and Design to build intuitive frontend interfaces (React, Next.js) for dashboards, configuration consoles, and visualization tools. Backend Services APIs:Develop and maintain the essential APIs (FastAPI/Python) and microservices that power the Vulcan platform. Architect robust task execution systems using Message Queues (Celery, RabbitMQ) to handle long-running asynchronous AI inference tasks. Internal Tooling Ecosystem: Build frontend and backend tools that accelerate internal teams (Project, ML/AI), including configuration panels, data visualization pipelines, and evaluation interfaces. Guardrails Security Features: Implement backend services for AI guardrails (content moderation, prompt filtering) and automated adversarial testing pipelines. 2. GenAI Agent Logic Engineering Prompt Engineering as Code: Treat prompts as software logic. Lead the design and implementation of AI Agent behaviors, optimizing responses through structured Prompt Engineering techniques. Agent Workflow Design: Orchestrate complex, multi-step LLM workflows where the "application logic" involves chaining model interactions effectively. Response Handling: Design robust parsing and validation mechanisms to ensure raw model outputs are converted into structured, usable application data. -

Our Product Vulcan is a cybersecurity solution specifically designed for GenAI, offering two core services: Red Team (vulnerability assessment) and Blue Team (real-time defense). It ensures GenAI compliance, cybersecurity robustness, and operational integrity. Since its official launch in 2024, Vulcan has been recognized by the international standard-setting organization OWASP as a certified vendor for LLM GenAI security testing and assessment. It is one of the few solutions capable of supporting multiple Asian languages (Traditional Chinese, Simplified Chinese, Japanese, Korean, Thai) and Standard Arabic. Learn more about us 👉 Vulcan product: https://vulcanlab.ai/Vulcan LinkedIn: https://www.linkedin.com/company/vulcanlab-ai/AIFT group: https://aift.io/Tech Blog: https://medium.com/onedegree-tech-blog About the role We are looking for a talented Machine Learning Engineer to join our Product Core Engineering team. You will be responsible for building and optimizing machine learning workflows that directly power our AI-driven products. This role focuses on the full lifecycle of model development — from training and fine-tuning to deployment and monitoring — ensuring robust and efficient ML systems at scale. Why Join Us? Product Impact: Your work will be directly embedded in our core AI products, shaping user experience and product capabilities.Engineering Excellence: Be part of a team that values high-quality engineering, reproducibility, and scalability.Innovation: Opportunity to experiment with cutting-edge ML and GenAI technologies in production settings.Collaboration: Work alongside backend, platform, and product teams in a highly collaborative environment.Competitive Package: Receive attractive compensation and benefits aligned with your skills and performance.How to Apply Please apply this position through 👉 https://job-boards.greenhouse.io/aift/jobs/5652688004 It will help us process your applications faster!! Key Responsibilities Model Development: Design and implement training processes for machine learning classifiers and generative models.Fine-tuning Prompting: Adapt pre-trained models to specific product needs through fine-tuning, prompt engineering, and parameter optimization.Hyperparameter Management: Configure and tune hyperparameters to balance accuracy, robustness, and performance.Pipeline Engineering: Build scalable training and evaluation pipelines to support continuous experimentation.Integration: Collaborate with backend and product engineers to deploy models into production systems.Monitoring Maintenance: Establish monitoring metrics and retraining strategies to maintain model performance in dynamic environments. -