Cake Job Search

【工作內容】職務說明主要工作內容: 1. 穩定生成人物影片以及指定動作行為等表現。2. 專注於研究、評估和整合最新的 AI 影像/影片生成模型 (如 Stable Diffusion、Wan 等) 的應用與技術 。3. 精通 Prompt Engineering，規劃和優化設計邏輯 ，以確保生成結果的品質與穩定性 。 【 工作待遇 】 面議，經常性薪資4萬以上【 必要條件 】．熟悉平台工具： ComfyUI 操作工作流。．熟悉AI 影像模型： Stable Diffusion, ControlNet, VAE, LoRA 。【加分條件 】 精通 Stable Diffusion 等主流 AI 影像生成模型的原理和參數配置。 具備圖像/影片生成 模型的訓練、微調（Fine-tuning）及部署經驗。 ．熟悉 Prompt Engineering 技巧，能透過文字精確控制生成結果 。．具備 LLM (如 GPT-4 、LLaMA3 ) 與影像模型整合的實戰經驗 。 ．具備跨模態系統架構思維與模組化開發能力 。

1. AI Agents 的導入： 　Ａ與各單位的使用者合作，理解單位需求，進行流程梳理、流程方案規劃與建議 　Ｂ負責AI Agents 與其他系統的整合 - 建立AI Agnets ，並且導入於使用者作業流程中 　Ｃ追蹤使用狀況、問題排除以及持續優化　Ｄ反饋問題與建議給AI Agents 平台 2. Domain 相關資料的資料治理： 　Ａ將資料按品質、用途做分層，讓資料的處理流程易於維護、追蹤與擴充

※請附上作品集或是作品連結網址。 【工作與職務內容】 ● 能夠與團隊成員、跨部門協作產出。 ● 能夠有效理解遊戲流程並轉換成影像。 ● 因應專案的特效需求，配合技術研究與功能開發。 ● 自製特效素材，具有影像合成、造型、具基礎手繪能力者佳。 【特質條件】 ● 熟悉CocosCreator遊戲引擎。 ● 熟悉AfterEffect、Spine任一套軟體。 ● 觀察力佳，有好的動態節奏感、物理反饋的理解。 ● 能獨當一面，能掌握專案特效的風格與品質。 ● 對特效有高度熱忱，對出現新技術、新功能會主動去研究了解。 ● 解決問題導向，針對跨部門溝通、產品體驗與表現都能主動思考更多層面的實現方式。 【加分條件】 ● 具備3D、TA相關作品或開發經驗。 ● 對產品有熱忱、願意深入理解玩法者。 ● 擅長AI技術並能實際應用在優化美術流程上。

Established in 1987 and headquartered in Taiwan, TSMC pioneered the pure-play foundry business model with an exclusive focus on manufacturing its customers’ products. As of 2024, TSMC serves more than 500 customers and manufactures over 11,000 products for high-performance computing, smartphones, the Internet of Things (IoT), automotive, and digital consumer electronics. It is the world’s largest provider of logic ICs, with an annual capacity of 16 million 12-inch equivalent wafers. TSMC operates fabs in Taiwan as well as manufacturing subsidiaries in Washington State, Japan and China, and the Company began construction on a specialty technology fab in Dresden, Germany, in 2024. In Arizona, TSMC is building three fabs, with the first starting 4nm production in 2025, the second by 2028, and the third by the end of the decade.Responsibilities: AI algorithm/application research and development on image recognition, time series forecasting, and abnormal detectionAI system design and development, including system infrastructure (docker/K8S) and web application.AI System analysis, design, development and integration for cross organizations/systems projects.Innovate, develop and introduce Intelligent Manufacturing Solutions.Hiring Organization: IMC

Established in 1987 and headquartered in Taiwan, TSMC pioneered the pure-play foundry business model with an exclusive focus on manufacturing its customers’ products. As of 2024, TSMC serves more than 500 customers and manufactures over 11,000 products for high-performance computing, smartphones, the Internet of Things (IoT), automotive, and digital consumer electronics. It is the world’s largest provider of logic ICs, with an annual capacity of 16 million 12-inch equivalent wafers. TSMC operates fabs in Taiwan as well as manufacturing subsidiaries in Washington State, Japan and China, and the Company began construction on a specialty technology fab in Dresden, Germany, in 2024. In Arizona, TSMC is building three fabs, with the first starting 4nm production in 2025, the second by 2028, and the third by the end of the decade.Advanced Quality System Development (AQSD) is part of TSMC's Intelligent Manufacturing Center (IMC), responsible for three main directions: Use statistical methods and conditional monitoring to inspect fab control charts, providing early warning of abnormal product yields; offer a fab defense system settings comparison platform to check that machines meet the internal control anomaly conditions set by the engineering department.Develop online fab analysis platforms through Big Data, Machine Learning, and Deep Learning techniques for online yield/defect analysis to ensure the quality of wafers produced in the fab.Develop systems and provide logical operations to control and implement escape mechanisms for material batch changes, measurement stages, and inspection stages, ensuring quality and optimizing product sampling rules. Responsibilities: Apply machine learning/optimization algorithm to build up models for optimizing semiconductor production.Develop analysis and optimization methods to enhance product quality, increase tool productivity and improve people productivity.Design, develop and test prediction models with real applications in manufacturing.Build a flexible framework to speed up the development process of AI models, with a focus on Large Language Models (LLMs) serving.Turn exciting AI prototypes/ideas into products, leveraging LLMs and other advanced AI technologies.Develop next-generation AI backend systems related to large-scale real-time data access, collection, analytics and monitoring.Establish and maintain MLops processes and tools, including model deployment, monitoring, and automation.Continuously improve the quality of AI production systems, particularly those utilizing LLMs.

建構與優化機器學習模型: 運用多維度運動數據（如賽事數據、球員表現），開發球員薪資計算、表現預測等核心模型，並持續進行優化。 支援產品與行銷決策: 深入分析用戶行為與市場趨勢，提供數據洞察，協助產品團隊優化功能、支援行銷團隊制定精準策略。 整合生成式 AI 應用: 導入並開發生成式 AI 功能，例如自動化產出運動新聞與球員動態，豐富產品內容。 提升內部數據工作流程效率: 運用 AI 技術優化資料處理、分析與建模的流程，提升團隊整體開發效率。

Established in 1987 and headquartered in Taiwan, TSMC pioneered the pure-play foundry business model with an exclusive focus on manufacturing its customers’ products. As of 2024, TSMC serves more than 500 customers and manufactures over 11,000 products for high-performance computing, smartphones, the Internet of Things (IoT), automotive, and digital consumer electronics. It is the world’s largest provider of logic ICs, with an annual capacity of 16 million 12-inch equivalent wafers. TSMC operates fabs in Taiwan as well as manufacturing subsidiaries in Washington State, Japan and China, and the Company began construction on a specialty technology fab in Dresden, Germany, in 2024. In Arizona, TSMC is building three fabs, with the first starting 4nm production in 2025, the second by 2028, and the third by the end of the decade. As an AI/ML Engineer, you will be a key member of a versatile team focusing on both algorithm development and platform implementation for TSMC’s RD, Fab, Business, IT, and Security functions to improve productivity and work quality. The role involves developing AI/ML solutions, implementing scalable platforms to support AI productionization, and maintaining the lifecycle of machine learning models. Responsibility:Common Responsibilities1. Collaborate with cross-functional teams of data scientists, ML engineers, and software developers. 2. Drive the complete AI solution development stages from data collection, exploration, and feature engineering to model evaluation and error analysis.3. Perform model lifecycle management to ensure the validity of models in production. 4. Exercise sound judgment in determining if business problems can be addressed through AI/ML and communicate results effectively to domain users. 5. Operate and maintain scalable, highly available AI platforms with continuous measurement and quality improvement to meet SLO/SLAs. Differentiated Responsibilities1. AI/ML Solution Development(1) Develop AI/ML algorithms that optimize manufacturing processes, improving productivity and work quality. (2) Explain AI/ML models to internal customers and conduct error analysis.2. AI Platform Development (1) Design and implement full tech stack solutions, including data-related technologies and AI platforms.(2) Develop tools, automation, and microservices to extend platform capabilities. (3) Manage container-based workloads and process high volumes of data in environments like Kubernetes.Additional information for the job: Job Location: Hsinchu Site, Taipei Office (Experienced only)On-call needs: On-call 1 week every 3 months The complete interview process includes:1. Manager interview2. Hackerrank test3. On-site personality and English test(which could be replaced if you have script of officially English test) 4. HR interview5. Second manager interview (Optional assessment)6. Technical review (Optional assessment)

Established in 1987 and headquartered in Taiwan, TSMC pioneered the pure-play foundry business model with an exclusive focus on manufacturing its customers’ products. As of 2024, TSMC serves more than 500 customers and manufactures over 11,000 products for high-performance computing, smartphones, the Internet of Things (IoT), automotive, and digital consumer electronics. It is the world’s largest provider of logic ICs, with an annual capacity of 16 million 12-inch equivalent wafers. TSMC operates fabs in Taiwan as well as manufacturing subsidiaries in Washington State, Japan and China, and the Company began construction on a specialty technology fab in Dresden, Germany, in 2024. In Arizona, TSMC is building three fabs, with the first starting 4nm production in 2025, the second by 2028, and the third by the end of the decade. The Facility Computer Integrated Manufacturing Department (FDCIM) is part of the Intelligent Manufacturing Center (IMC) at TSMC. It is primarily responsible for the development and maintenance of the following products: Development and maintenance of manufacturing-related report products for wafer fabs.Development and maintenance of engineering-related report products for wafer fabs.Development and maintenance of digital operation system products for facility management.Development and maintenance of big data application system products for facility management.Development and maintenance of facility management AI and machine learning related algorithm development and application system products. Responsibilities: FDCIM also employs software engineering and modular development techniques, combined with high-performance database application technologies, to develop systematized software with a unified version control system that accommodates different time zones and languages globally.In addition to its regular software product development work, FDCIM is also engaged in the research and development of new technologies, including the application of DevOps, Microservices, MLOps, AIOps, and more. Develop and maintain AI/ML systems and algorithmsCollaborate with cross-functional teams to identify and solve business problems using AI/ML techniquesDesign and implement machine learning models and data pipelinesTest and validate AI models for accuracy, scalability, and efficiencyDeploy AI solutions to production environmentsStay up-to-date with the latest advancements in AI technologies and industry trendsWrite clean and efficient code using HTML, CSS, and JavaScriptOptimize web applications for speed and scalabilityTest and debug web applications across multiple browsers and devicesStay up-to-date with the latest front-end development trends and best practices

► 職務執掌作為前端工程師，你將負責網頁框架建構、元件開發與 API 串接，並與後端團隊協作，確保介面順利整合。透過精準的切版與互動效果實作，延續設計品質，讓理想的使用體驗真實落地。► 期待共創非凡我們期待你是樂於分享、勇於挑戰的夥伴，願將所學、所長與洞見帶入團隊，在合作中持續成長、共創價值。只要你具備積極自主的態度，就有機會成為推動團隊前進的關鍵力量。同時，我們希望你對 AI 技術保持好奇與開放，善用如 ChatGPT、OpenAI API、GitHub Copilot 等工具提升開發效能。公司提供相關工具補助與定期研討，支持每位夥伴精進專業、掌握未來開發趨勢。

Established in 1987 and headquartered in Taiwan, TSMC pioneered the pure-play foundry business model with an exclusive focus on manufacturing its customers’ products. As of 2024, TSMC serves more than 500 customers and manufactures over 11,000 products for high-performance computing, smartphones, the Internet of Things (IoT), automotive, and digital consumer electronics. It is the world’s largest provider of logic ICs, with an annual capacity of 16 million 12-inch equivalent wafers. TSMC operates fabs in Taiwan as well as manufacturing subsidiaries in Washington State, Japan and China, and the Company began construction on a specialty technology fab in Dresden, Germany, in 2024. In Arizona, TSMC is building three fabs, with the first starting 4nm production in 2025, the second by 2028, and the third by the end of the decade. The Intelligent Process Control Systems Department (IPCS) in TSMC's Intelligent Manufacturing Center (IMC) assists fabs in analyzing process issues. By employing techniques such as experimental design, statistical analysis, optimization algorithms, and machine learning, the department establishes R2R (Run-to-Run) process control models. Subsequently, in collaboration with the system development team, an automated control platform is established and rolled out across various fab zones. This integration aims to enhance staff efficiency and improve overall product quality. Responsibilities: By employing methods such as control system development, algorithm design, and control performance diagnostics, we provide advanced control systems for the semiconductor industry.We use machine learning and optimization algorithms to build models that deliver solutions meeting the needs of the fab zones.We establish automated platforms that utilize techniques such as machine learning and AI for automatic parameter optimization, which reduces staff workload and enhances quality.In collaboration with the system development team, we systematically implement and roll out innovative methods or improvements across various fab zones to boost staff efficiency and product quality.