「機器學習主任工程師 Principal Engineer（Audio）-E110707」的相似工作

We are seeking a talented and passionate Audio AI Engineer to join our team. You will be responsible for developing, training, and deploying cutting-edge AI models for real-time and audio noise reduction. Your work will directly impact the clarity and quality of our products, from teleconferencing and communication tools to creative audio production and media. The ideal candidate has a strong background in deep learning, a solid understanding of digital signal processing (DSP) fundamentals, and a keen ear for sound. You will collaborate with a cross-functional team of researchers, software engineers, and product managers to bring innovative audio solutions to life. # Responsibilities 1. Design, develop, and train deep learning models (e.g., neural networks) specifically for audio denoising, dereverberation, and speech enhancement. 2. Curate, manage, and preprocess large-scale audio datasets for model training and evaluation. This includes collecting, cleaning, and labeling noisy and clean audio pairs. 3. Optimize AI models for performance, latency, and resource efficiency, ensuring they can run effectively on various platforms, including cloud servers, mobile devices, and embedded systems. 4. Stay up-to-date with the latest research in audio signal processing, machine learning, and generative AI for audio. Explore and prototype new algorithms and techniques to push the boundaries of audio quality. 5. Work closely with the software engineering team to integrate production-ready models into our software and services. Partner with product and design teams to understand user needs and translate them into technical requirements. 6. Develop robust evaluation metrics and conduct rigorous testing (both objective and subjective listening tests) to ensure the denoising models meet high-quality standards. 7. Diagnose and resolve technical issues related to audio processing pipelines and AI model performance. # Requirements 1. Bachelor's, Master's, in Computer Science, Electrical Engineering, Signal Processing, or a related technical field. 2. Proven experience in a role focused on machine learning, with a specialization in audio, speech, or signal processing. 3. Strong proficiency in Python is a must. Experience with C++ is a plus, especially for low-latency, real-time applications. 4. Hands-on experience with deep learning frameworks such as PyTorch or TensorFlow. 5. A solid theoretical and practical understanding of digital signal processing (DSP) concepts, including time-frequency analysis, filter design, and spectral processing. 6. Familiarity with audio editing software like Audacity, Pro Tools, or Adobe Audition for data inspection and analysis. 7. Proficiency with Git. # Preferred Qualifications 1. Experience in a production environment deploying machine learning models into live systems. 2. Knowledge of specific audio denoising techniques and models (e.g., RNNoise, WebRTC, U-Net architectures for audio). 3. Experience with real-time audio systems and APIs. 4. A background in audio engineering or a related field, with an excellent ear for sound quality. 5. Strong mathematical foundation in linear algebra, probability, and statistics. 6. Experience with cloud platforms (AWS, GCP, Azure) and MLOps pipelines. # Soft Skills 1. Excellent problem-solving and analytical skills. 2. Strong communication and collaboration abilities, with the capacity to explain complex technical concepts to non-technical stakeholders. 3. A passion for audio and a creative mindset for solving difficult audio challenges. 4. Ability to work independently and as part of a highly collaborative team. 5. Proficiency in both English and Chinese, written and spoken.

請上傳英文履歷至官網(職缺代碼12427): https://careers.synopsys.com/job/hsinchu/r-and-d-engineering-sr-staff-engineer/44408/84643890752 You Are: You are a passionate and inventive engineer eager to advance the state of semiconductor technology. With a strong foundation in device physics, circuit simulation, and programming, you excel at translating complex technical challenges into innovative solutions. What You’ll Be Doing: 1.Researching, designing, and implementing advanced semiconductor component models, supporting both mainstream and emerging technologies. 2.Collaborating with peer R&D teams to integrate new technologies into SPICE/FastSPICE products, driving performance optimization. 3.Partnering with industry leaders to develop custom modeling and characterization methodologies tailored to cutting-edge applications. 4.Working closely with teams specializing in parasitic extraction, standard cell library characterization, and related domains to create comprehensive solutions for emerging technologies. 5.Providing expert-level technical support and guidance to key and strategic customers, ensuring their success with Synopsys products. 6.Driving innovation by contributing to the development of reliability modeling and advanced analysis features within SPICE/FastSPICE simulation tools. 7.Participating in cross-functional teams to ensure seamless product integration and a best-in-class user experience. The Impact You Will Have: Accelerate the adoption of next-generation semiconductor technologies through state-of-the-art modeling and simulation capabilities. Enhance the performance and reliability of SPICE/FastSPICE simulation products, directly influencing the success of global semiconductor leaders. Empower customers to innovate faster and with greater confidence by providing robust, accurate, and scalable simulation solutions. Shape the future direction of Synopsys’ EDA offerings by contributing to groundbreaking feature development and integration. Strengthen Synopsys’ reputation as a trusted partner and technology leader within the semiconductor industry. Drive industry standards for device modeling, reliability analysis, and advanced circuit simulation methodologies. Foster an inclusive and collaborative culture, mentoring junior engineers and promoting best practices across the organization. What You’ll Need: 1.MS or PhD in Electrical Engineering, Computer Science, Physics, Material Science, Applied Statistics, or Applied Mathematics, with 3+ years of relevant industry experience. 2.Strong knowledge of device physics, semiconductor process, and manufacturing flows. 3.Proficiency in programming with Linux, C/C++, and Python; solid foundation in data structures and algorithms. 4.Hands-on experience in SPICE circuit simulation or the development of related EDA software products. 5.Familiarity with Memory, Analog, Digital, High Frequency, or RF device/circuit design (a plus). 6.Experience or interest in machine learning applications and reliability analysis methodologies (a plus). Who You Are: Exceptional communicator, able to explain technical concepts clearly to diverse audiences. Strong problem-solver with a can-do attitude and resilience in the face of technical challenges. Team player who thrives in collaborative, multicultural environments. Self-motivated, organized, and able to manage multiple priorities effectively. The Team You’ll Be A Part Of: You’ll join a world-class R&D team at the forefront of SPICE/FastSPICE circuit simulation technology. The team is dedicated to developing, innovating, and implementing new technologies for device and reliability modeling, as well as advanced analysis features. Working closely with other market-leading product teams and industry partners, you will contribute to cross-functional projects that drive Synopsys’ leadership in EDA solutions. Our team values collaboration, knowledge-sharing, and a commitment to excellence in all aspects of product development and customer support.

CerenceAI China RD is seeking a Research Engineer to design and implement the next-generation text-to-speech systems and applications. In this role, with members in globe, you will work both frontend (Grapheme-to-phoneme, Text Normalization, phrasing and prosodic control, etc) and backend (acoustic modeling, neural vocoding) components of engine pipelines and end-2-end system for major languages and dialects. Your work will directly impact CerenceAI products ranging from voice assistants to accessibility tools, delivering natural, expressive, and multilingual speech synthesis. Job Description Representative responsibilities/duties will include but not limited to: • Design and optimize text/NLP preprocessing pipelines with Deep Learning or Machine Learning methods, including Grapheme-to-phoneme (G2P) conversion for multilingual support; Text normalization; polyphone disambiguation; Prosody prediction and control • Integrate language models (e.g., BERT, GPT variants) to improve contextual and semantic understanding for natural intonation • Develop rule-based and neural solutions for emotion/style control in synthesized speech • Build state-of-the-art acoustic models (e.g., Tacotron, FastSpeech, VITS) to map linguistic features to spectrograms or waveform parameters. • Optimize neural vocoders (e.g., WaveNet, HiFi-GAN, MelGAN, LPCNet) for high-fidelity, real-time speech synthesis • Optimize inference latencies for both edge devices and cloud platforms • Enhance robustness through noise suppression, speaker adaptation, and multilingual/cross-language/cross -gender voice cloning Knowledge, skills, and qualifications Education: Master in CS, AI, EE, Math, or related field. Required/preferred skills: • hands-on experience in speech generative system development with deep expertise in both frontend and backend components • Proficiency in C/C++ and Python, with mastery of ML frameworks (PyTorch, TensorFlow, etc) • Some background in NLP techniques and/or speech signal processing is welcome • Knowledge on transformer-based language models for prosody prediction • Basic understanding of autoregressive / non-autoregressive acoustic models and neural vocoders • Experience in optimizing models via quantization, pruning, or knowledge distillation • Experience with ONNX Runtime, TensorRT, or TorchScript, etc • Experience with zero-shot/one-shot/few-shot voice cloning or emotional TTS systems • Skilled GPU/TPU cluster and grid user • Fluent English is a must-have

CerenceAI China RD is seeking a Research Engineer to design and implement the next-generation text-to-speech systems and applications. In this role, with members in globe, you will work both frontend (Grapheme-to-phoneme, Text Normalization, phrasing and prosodic control, etc) and backend (acoustic modeling, neural vocoding) components of engine pipelines and end-2-end system for major languages and dialects. Your work will directly impact CerenceAI products ranging from voice assistants to accessibility tools, delivering natural, expressive, and multilingual speech synthesis. Job Description Representative responsibilities/duties will include but not limited to: • Design and optimize text/NLP preprocessing pipelines with Deep Learning or Machine Learning methods, including Grapheme-to-phoneme (G2P) conversion for multilingual support; Text normalization; polyphone disambiguation; Prosody prediction and control • Integrate language models (e.g., BERT, GPT variants) to improve contextual and semantic understanding for natural intonation • Develop rule-based and neural solutions for emotion/style control in synthesized speech • Build state-of-the-art acoustic models (e.g., Tacotron, FastSpeech, VITS) to map linguistic features to spectrograms or waveform parameters. • Optimize neural vocoders (e.g., WaveNet, HiFi-GAN, MelGAN, LPCNet) for high-fidelity, real-time speech synthesis • Optimize inference latencies for both edge devices and cloud platforms • Enhance robustness through noise suppression, speaker adaptation, and multilingual/cross-language/cross -gender voice cloning Knowledge, skills, and qualifications Education: Master in CS, AI, EE, Math, or related field. Required/preferred skills: • hands-on experience in speech generative system development with deep expertise in both frontend and backend components • Proficiency in C/C++ and Python, with mastery of ML frameworks (PyTorch, TensorFlow, etc) • Some background in NLP techniques and/or speech signal processing is welcome • Knowledge on transformer-based language models for prosody prediction • Basic understanding of autoregressive / non-autoregressive acoustic models and neural vocoders • Experience in optimizing models via quantization, pruning, or knowledge distillation • Experience with ONNX Runtime, TensorRT, or TorchScript, etc • Experience with zero-shot/one-shot/few-shot voice cloning or emotional TTS systems • Skilled GPU/TPU cluster and grid user • Fluent English is a must-have

我們是一家專注於AI工程的新創公司，正在尋找 AI 工程師，負責開發以 AI Agent 技術提升專業領域工作效率。 核心職責： ．AI Agent 開發與應用：設計並訓練 AI Agent，提升專業領域效率與準確性。 ．知識圖譜建構：整合專業領域知識，建立結構化知識圖譜，支援 AI Agent 進行推理與決策。 ．生成式 AI 應用：利用 LLM、多模態模型輔助、優化與驗證。 ．最佳化與強化學習：應用 GA(遺傳演算法)、RL(強化學習)等技術，開發與優化演算法。 ．跨領域協作：與專業領域合作，驗證 AI Agent 產出的結果，確保其符合標準與規範。 必要條件 ．熟悉機器學習、深度學習相關技術 (PyTorch / TensorFlow / JAX)。 ．具備 AI Agent 相關開發經驗，如 LLM + Tool-Use、強化學習或知識圖譜應用。 ．熟悉數值最佳化、啟發式演算法 (GA、PSO、Bayesian Optimization 等)。 ．具備 Python 開發經驗，並熟悉 AI/ML 相關框架與工程應用。 我們團隊致力於最前沿 AI Agent 技術應用，挑戰與業界專家深度合作，快速累積專業領域 AI 技術經驗。歡迎對 AI有熱情的你加入！

1. 負責鋰電池模組BMS系統規畫、硬體電路設計、設計審查與測試驗證。 2. 負責電池模組BMS新平台規劃設計、技術開發與專利佈局。 3. 負責監督專案執行、制定硬體設計規範、解決方案、開發品質管理和流程優化。 4. 負責硬體開發成本控制。 5. 負責管理硬體部門、組織規劃、人才招募。 6. 主管交辦事項。

[General Summary] As a forward-thinking technology company, Augentix advances the limits of innovation in "Industrial and Embedded IoT" to deliver next-generation experiences and accelerate digital transformation toward a smarter, more seamlessly connected world. As a AI & Machine Learning Software Engineer at Augentix, you will play a critical role in enabling high-performance, efficient, and secure AI solutions across embedded and edge platforms. You will contribute to the full development lifecycle of our AI software stack, from model conversion and optimization to deployment and system integration. Working closely with cross-functional teams, you will help deliver scalable AI pipelines and frameworks that power advanced imaging, sensor fusion, and industrial intelligence use cases in real-world environments. [Responsibilities] ★ Implement AI model conversion, quantization, optimization, and secure deployment. ★ Optimize performance, power, and memory bandwidth in AI/DSP pipelines. ★ Analyze and debug AI accuracy and performance issues reported by internal teams or during system-level integration. ★ Design and develop AI-based software frameworks targeting Linux and embedded platforms. ★ Collaborate with hardware, systems, and test teams to deliver robust end-to-end solutions. ★ Occasional business travel across APAC and other regions may be required. [Minimum Qualifications] ★ Master's degree in Electrical Engineering, Computer Science, Mathematics, Physics, or a closely related field with 3+ years of relevant experience, or a PhD in a related field. ★ Proficient in modern C, C++, and Python. ★ Experience in neural network deployment, quantization, and model compression. ★ Solid understanding of neural network inference frameworks for embedded systems (e.g., TFLite, TFLite Micro, NCNN). ★ Good understanding of neural network training frameworks (e.g., TensorFlow, PyTorch). [Preferred Qualifications] ★ Familiarity with hardware accelerators and embedded DSP architectures. ★ Experience in video/image processing, computer vision, or multimedia algorithm development. ★ Experience with GStreamer is a plus. ★ Experience with NN SDKs and tools, specifically "AI Engine Direct SDK (QNN)" and "AI Model Efficiency Toolkit (AIMET)", is a strong plus. ★ Comfortable working in a globally distributed, cross-disciplinary engineering team.

We are seeking a highly skilled ML Engineer with expertise in Natural Language Processing (NLP), Transformers, and Large Language Models (LLMs) to join our team. This role will focus on building and optimizing scalable model training infrastructure in cloud environments, with an emphasis on cloud technologies, cloud-based GPUs, and MLOps pipelines. The ideal candidate will have a strong foundation in developing and deploying ML models, particularly in the LLM domain, and experience with cutting-edge frameworks and cloud-native tools. Responsibilities: - Design, implement, and optimize scalable machine learning training pipelines tailored for Transformers and LLM-related technologies. - Fine-tune and deploy state-of-the-art models, including transformers and LLMs (e.g., GPT, Llama, NVIDIA NeMo..etc), for various cloud-based MLOps pipelines. - Utilize cloud-based GPU resources for efficient model training and experimentation. - Collaborate with cloud engineers to design robust MLOps pipelines and deploy trained models into production environments. - Monitor, maintain, and optimize model performance in production, ensuring reliability and scalability. - Research and integrate the latest advancements in related technologies into model training workflows. - Contribute to the continuous improvement of the ML infrastructure, focusing on automation and reproducibility. Qualifications: - Bachelor’s or Master’s degree in Computer Science, Machine Learning, Data Science, or a related field. - Strong understanding of machine learning techniques, with a focus on deep learning, NLP, and transformer-based architectures. - Proficiency in PyTorch or TensorFlow, with hands-on experience in fine-tuning AI models. - Experience in designing and executing machine learning model training workflows, including data preparation, feature engineering, hyper-parameter tuning, and performance evaluation. - Familiarity with MLOps principles, tools, and frameworks, including CI/CD pipelines for ML models. - Strong programming skills, particularly in Python (required), and familiarity with other languages like Java or other languages is a plus. - Excellent problem-solving, analytical, and communication skills.

AI is currently at the forefront of technological advancement, marking a transformative shift across industries. As a global cybersecurity leader, Trend Micro is committed to staying at the cutting edge of this evolution, with the goal of integrating AI technology to make the exchange of digital information safer and more secure. Fueled by decades of security expertise, world-leading global threat research and intelligence, and continuous innovation, our cybersecurity platform protects hundreds of thousands of organizations and millions of individuals across clouds, networks, devices, and endpoints. The objective of TW Engineering Group 1 is to deliver the best endpoint security solutions that help commercial customers face cybersecurity challenges effectively. We are dedicated to protecting businesses of all sizes—from small and medium-sized companies to large enterprises—so they can operate securely and confidently in today’s digital world. At the same time, we continue to provide security, privacy, and anti-scam solutions for the consumer market, using our decades of experience to protect millions of individual users worldwide. By combining insights from both commercial and consumer products, we aim to create innovative solutions that solve real-world problems and keep all of our customers safe. Everyone in Group 1 has the opportunity to join different teams and share their knowledge and passion, contributing to real results for our customers. Along this journey, we look for people who not only bring strong technical skills but also a forward-thinking mindset—those who can leverage AI technology to challenge the status quo, enhance productivity, and build innovative solutions that truly make a difference. We believe in adopting AI-powered development workflows to help our engineers produce more reliable, and high-quality programs, and create scalable products faster. This helps us learn from our customers and build solutions that truly make a difference. As a (Sr.) Software Engineer, you will be responsible for designing, developing, testing, deploying, and supporting high-quality software solutions. This role involves close collaboration with architects, project managers, developers, QA, and UI designers to develop products that satisfy customer needs, solve their problems, and achieve customer success. [Your Mission] 1.Leverage AI-assisted development tools (e.g., GitHub Copilot, ChatGPT or MCP servers) to boost development speed, enhance code quality, and accelerate prototyping. 2.Collaborate closely with stakeholders on requirement discussions, cross-team integration, and issue resolution. 3.Deliver features through the full development lifecycle, including planning, design, implementation, testing, deployment, and operation. 4.Create and execute test cases and scripts to validate software functionality, performance, reliability, and security. 5.Troubleshoot and debug complex issues by identifying root causes and implementing effective solutions. 6.Automate and enhance development and release processes to improve efficiency and reliability.

We are seeking a highly skilled **Embodied AI Model Security Engineer** to join our AI research and robotics engineering team. This hybrid role blends **AI/ML expertise** with **robotics simulation and control**, focusing on the development, fine-tuning, and secure deployment of embodied AI systems. You will design training pipelines, work with robot simulation environments, implement teleoperation data collection frameworks, and research model security to safeguard our AI-driven robotics applications. ## Key Responsibilities ### AI/ML Development & Optimization * Design and implement **end-to-end fine-tuning pipelines** for large-scale language, vision, and vision-language-action (VLA) models. * Develop and manage **training datasets**, including synthetic data generation from simulators and real-world data. * Profile, benchmark, and optimize inference **latency, memory footprint, and throughput** for both cloud and edge devices. * Research and analyze AI model vulnerabilities and develop defense methods. ### Robotics Simulation & Control * Build and operate robotics simulation environments using **IsaacSim, Gazebo, MuJoCo**, or similar tools, and operate physical robots via robotics operating systems such as ROS/ROS2. * Integrate AI models into simulation pipelines for testing, evaluation, and synthetic data generation. * Validate control policies in simulation before deployment to physical hardware. ### Robot Communication & Control Systems * Implement robot control via **CANBus, serial, or similar communication protocols**. * Develop real-time interfaces between AI models and robot hardware controllers. ### Leader–Follower Teleoperation for Data Collection * Implement **leader–follower architectures** to teleoperate robots for **training data collection**. * Capture and preprocess trajectories for use in imitation learning or reinforcement learning pipelines. ### AI Security Research & Defense * Independently research AI model vulnerabilities, focusing on embodied AI and robotics applications. * Analyze **adversarial attacks, model extraction, and unsafe behaviors**, and design mitigation strategies. * Document security findings and propose architectural improvements for safe model deployment. --- ## Minimum Qualifications * Master’s or PhD in **Computer Science, Machine Learning, Robotics**, or a related field. * **5+ years of AI/ML engineering experience**, with proven expertise in fine-tuning and deploying large-scale transformer or VLA models. * Proficiency in **Python** and major AI frameworks such as **PyTorch, TensorFlow, or JAX**. * Experience working in **Linux/Unix** environments. * Hands-on experience with **robotics simulators** (IsaacSim, Gazebo, MuJoCo, etc.). * Familiarity with **CANBus, serial communication**, or other robotics control interfaces. --- ## Preferred Skills & Experience * Experience with **adversarial machine learning** or security evaluation of AI models. * Previous work with **LeRobot** or similar open-source robotics frameworks, including controlling robots via VLA models. * Experience integrating AI models with **Kubernetes, Docker, or NVIDIA TensorRT** for deployment. * Familiarity with **differential privacy, secure enclaves, and threat modeling** for AI systems. * Track record of **technical publications or patents** in AI, robotics, or security.

1.音訊演算法開發 2.算法驗證調適與優化 3.客戶問題分析及除錯

產品專頁網址- https://www.solomon-3d.com/tw/home-tw/ 產品介紹影片- https://www.solomon-3d.com/tw/solomon-meta-aivi/ 職務內容： 1. 推廣AI視覺與3D辨識產品，進行業務簡報與應用建議 2. 客戶關係維護和開發新客戶 3. 達成營運績效和市場資訊蒐集回饋 其他條件： 1. 具備自動化設備、工業系統整合或機器視覺產品銷售經驗 2. 對AI與智慧製造有實際接觸或濃厚興趣 3. 喜歡與客戶互動，有解決問題的思維與責任感 4. 想在技術型業務領域長期發展

本公司核心為AI技術解決方案，改變人們的生活品質，目前致力研發專注於智慧助聽技術與人工智慧視覺應用類別的產品。 【工作內容】 1. 結合深度學習（DL）與傳統電腦視覺，研發影像強化、物件偵測、物件分割等相關技術，並實作於 Edge Device。 2. . 完成主管交辦事項。 【職務需求】 1. 具備人工智慧與電腦視覺開發經驗。 2. 熟悉 C/C++ （尤佳）。 3. 熟悉 OpenCV。 4. 至少擅長一項深度學習框架 （如：Pytorch、Tensorflow）。 5. 具 Git 版本控制知識。 6. 具獨立開發能力。 7. 具良好團隊合作能力，有意願與團隊夥伴分享技術知識。 【加分項目】 1. 具 Embedded System 系統整合經驗。 2. 有布署 AI 模型至嵌入式系統經驗。 3. 熟悉 AI 模型剪技、量化、蒸餾等。

Immersive Optics System Engineer / AR VR Engineer ⭐有機會達到百萬年薪!歡迎對外商感興趣的你，保障年薪13.5個月⭐ Responsibilities: Provides expert knowledge regarding the selection, evaluation, qualification and procurement of visual system in different varieties of display size applications. Collaborates and communicates with internal and outsourced development partners about design requirements, regulatory requirements, and user needs. Participates as a member of project team of other engineers, Program Managers, and internal and outsourced development partners to develop reliable, cost effective and high-quality solutions for the immersive visual system products. Experience on NPI and manufacture process. Familiar with display industry production. EE experience to bring up the display system (including MIPI / Vx1 / eDP etc…). To review the EE design with supplier and system designer. Knowledge of HDMI and DP. Education and Experience Required: MS in Display Engineering, Materials Science, Optical Engineering, Physics or related field and/or equivalent experience. Typically 5+ years experience, with 3+ years at related field. Knowledge and Skills: Solid knowledge in different form factor product. Such as VR/AR, Monitor, TV, AIO etc. Display design (Array / FEOL / BEOL) and quality control Display system hardware design. Basic knowledge of Geometrical optics and Optical imagery. Camera ALS / Proximity Sensor Depth sensing (Structure light/TOF) Good communication skills in English and Mandarin with ability to work under pressure. Proven track of record on development and implementation of innovative LCD, AMOLED, or other display solutions for consumer electronics. Experience with at least one of the focus areas of visual system technology below: -VR display system (Panel + Lens) -Video passthrough system (Dual camera + ISP) -AIO touch monitor system -IWB system -SLAM system (camera + sensor fusion)

★工作內容 1.設計、開發及維護生成式AI應用及流程。 2.以演算法及技術調校及精進生成式 AI 模型。 3.協助研發軟體新技術與新工具。 4.評估客戶需求，協助AI落地應用。 5.LLM(RAG)應用開發。 6.微調和優化LLM模型。 7.持續關注LLM、NLP與生成式AI技術的最新進展並將其應用於公司的項目。 8.確保LLM模型在生產環境中的可擴展性和效率。 9.大語言模型(LLM)訓練之資料整備與標註。 ★必備條件 1.具有Python 3軟體開發二年以上之實務經驗。 2.熟悉使用Docker, Git, Github等工具。 3.了解大模型基礎知識及常見微調手段，熟悉Mistral、LLama等模型微調訓練、優化方案。 4.具有Linux環境開發經驗。 5.熟悉Pytorch、Huggingface等深度學習工具。 6.具有機器學習/深度學習/生成式AI相關研究或程式開發經驗。 7.熟悉NLP和RAG，有實務經驗。 8.具備LLM模型微調經驗，熟悉Llama 3+、Breeze、OpenAI和Gemini等模型。 ★加分條件 1.具利用LangChain/LlamaIndex的開發經驗。 2.在文案生成、chatbi、智能客服、個人助手等至少一個業務場景有專案經驗。 3.有過Chatbot 、生成式AI應用相關工作經驗。 4.具利用NVIDIA NeMo（NIM）的開發經驗。 5.熟悉ChatGPT等大語言模型應用。 6.ChatGPT應用經驗：有使用ChatGPT或相關大型語言模型進行專案開發的經驗，了解其7.在各種應用場景中的潛在用途。 8.熟練提示調整技術（Prompt Tuning），善用Prompt來改善AI模型的回答效果。 9.能分析不同Prompt對模型表現的影響，並進行持續的調整和改進。 10.熟悉Diffusion Model、Transformer Model等生成式AI模型。 11.熟悉Computer Vision相關深度學習模型實作與pytorch或tensorflow等框架。 12.熟悉C/C++。 13.有AI演算法相關論文研究經驗。 14.熟悉TS，深度了解ML、LLM、RAG、Finetune、Agent原理。

Embedded Vision Engineer About the Role We are seeking a skilled Embedded Vision Engineer to join our "Special Projects" team. The ideal candidate will have hands-on experience with NVIDIA Jetson platforms and a strong proficiency in OpenCV. In this role, you will be responsible for developing and optimizing computer vision applications for edge devices, integrating sensor data, and deploying efficient, real-time algorithms on our industrial embedded platforms. You will report directly to the Chief Technology Officer (CTO). About Foreman Technologies Foreman Technologies develops advanced autonomous robotic solutions for a variety of industrial and commercial applications, including inspection and painting of commercial buildings, as well as marine and petroleum infrastructure. Our headquarters are located in the United States, with regional offices in Arizona, Texas, and Virginia. This position is based at our R&D center in New Taipei City, Taiwan. Key Responsibilities AI Camera Application Development Design, develop, and deploy robust computer vision applications with a focus on real-time solutions for security and surveillance. Implement and optimize core functionalities, including real-time object detection, intrusion detection, and person tracking. Utilize and integrate core computer vision and machine learning libraries such as OpenCV, TensorFlow, and PyTorch. Apply advanced algorithms like YOLO for efficient object detection on edge computing platforms. Embedded Systems & Hardware Integration Develop on NVIDIA Jetson devices (e.g., Xavier, Nano, Orin) and other embedded edge computing platforms. Integrate and process data from various camera modules and sensor inputs, with a strong understanding of protocols like ONVIF and RTSP. Optimize the performance of computer vision applications using hardware acceleration techniques such as CUDA, TensorRT, or DeepStream SDK. System Integration & Collaboration Develop interfaces to connect vision components with higher-level application logic in languages such as Python or C++. Collaborate closely with hardware, machine learning, and robotics teams to seamlessly integrate solutions into real-world systems. Conduct thorough testing, debugging, and profiling of embedded vision applications to ensure performance and stability. Documentation & Maintenance Maintain comprehensive technical documentation. Provide technical support during product deployment and field testing. Qualifications Required Skills & Experience Proven experience in computer vision, including object detection, person tracking (and trajectory analysis), and facial recognition. Hands-on experience with deploying and optimizing machine learning models using tools like ONNX. Experience with AI camera development is highly preferred. Additional Requirements Nationality & Residency: This position is reserved for Taiwanese nationals with household registration in Taiwan. We are unable to sponsor work visas for international candidates. Language Proficiency: Native-level fluency in Mandarin is required. Exceptional verbal and written English proficiency is essential for frequent communication with our US-based leadership team. Interviews may be conducted partially or entirely in English.

1.AI / 深度學習基礎 　熟悉 Python 與主流 AI 框架（PyTorch / TensorFlow）。 　具備 深度學習模型訓練與推論 經驗（特別是語音與 NLP 領域）。 　理解 大型語言模型（LLM） 基本原理與應用方式。 2.語音相關技術 　熟悉 語音轉文字（ASR） 技術，具備 Whisper 或其他 STT 模型使用經驗。 　了解 聲學模型、語言模型、分詞與斷句處理。 　有 多語言語音處理 經驗（中文、台語、英語尤佳）。 　了解 文字轉語音（TTS） 框架，如 Tacotron、VITS、Parler-TTS 等。 3.NLP 與翻譯 　具備 NLP 模型 的應用經驗（如 Hugging Face Transformers）。 　了解 機器翻譯（MT） 技術，能調整/微調翻譯模型。 　對 語意理解、摘要、對話系統 有基礎知識。 4.AI Edge / 系統開發 　熟悉 AI Edge 裝置開發（NVIDIA Jetson、Raspberry Pi、ARM 架構等）。 　具備 模型壓縮、量化、部署 經驗（TensorRT、ONNX Runtime、OpenVINO）。 　能處理 效能優化、低延遲推論 與裝置資源限制。 　熟悉 Linux 環境與 Docker/Kubernetes。 5.後端與系統整合 　熟悉 FastAPI / Flask / gRPC 等 API 框架。 　具備 SQL / NoSQL 資料庫應用經驗。 　熟悉 資料管線工具（Dagster / Airflow）。 　有 MLflow 或其他 MLOps 工具 使用經驗。 6.加分條件 　有 即時語音應用（會議記錄、字幕、智慧助理）開發經驗。 　熟悉 多語言、多口音語音辨識挑戰。 　有 演算法優化 或 自研語音模型 經驗。 　熟悉 前端/行動裝置整合（如 React、Flutter）更佳。

1. 負責自然語言處理、機器學習/深度學習模型相關模型的設計與應用，包含訓練、參數調整與部署。 2. 整合開源框架，開發與實作具任務導向的 AI Agent 系統，結合語言模型與外部工具自動完成任務。 3. 分析模型與代理人執行成效分析，進行回訓、評估與優化。 4. 探索技術新知，研讀最新 AI 論文或開源工具，並進行內部實驗與導入評估。 5. 撰寫清晰的技術文件與報告，支援跨部門合作與知識傳承。

【職務內容】 你將參與我們 AI 應用產品的設計與開發，包含： －AI 模型、相關服務評估與整合 －API 介面設計與開發 －設計資料處理流程，確保 AI 功能整合順暢 －系統維護及問題修正 －積極追蹤並導入 AI 新知，定期與團隊分享或技術導入 【Must Have】 －熟悉 Python、關聯式資料庫、SQL等相關技術，以及當前生成式模型發展 －有實務整合過 OpenAI API 或其他 LLM －熟悉 API 設計，能獨立設計系統架構並處理與外部服務串接 －用過 RAG、embedding、向量資料庫...等常用技術或工具 －具備 prompt 設計經驗，了解如何優化 prompt 以符合特定任務 【Nice to Have】 －具備主動學習新技術心態 －若具備電商領域背景，將更能理解用戶場景並優化 AI 應用策略

【團隊介紹】 91APP是台灣首家掛牌上櫃的 SaaS 軟體服務公司。91APP 為國內第一家「虛實融合 (Online Merge Offline, OMO)」新零售軟體雲服務公司，結合虛實整合並提供線上線下管理工具，協助零售品牌快速切入電商市場，整合線上線下雙向導流、銷售訂單及全通路會員經營數據，快速增強營運效率提升獲利。 在 91APP 的資料科學團隊，我們擁有的是研究不完的案例，你可以了解每一家品牌是如何從剛起步擴展到月業績破千萬，你可以仔細研究並探討其每個階段的關鍵原因，然而這些有趣的小結果都將內化堆疊成為你的產業知識，或是變成標準軟體產品提供給全台灣甚至全世界的品牌使用。 在這裡，你有機會站在第一線直接面對世界級的品牌對他們講述你從數據中發現精彩萬分的商業故事；你也可能在產品團隊中跟你的產品經理、使用者經驗設計師與開發團隊們討論你苦心研究多時的數據模型需要如何對市場溝通、要設計什麼流程來吸引更多品牌使用以及要如何將你的模型部署至正式環境。這些內容將讓你不僅保有商業現場的敏銳度也同時具有軟體開發的技術力。當然，無論你想往前（商業現場）多跨一步或是往後（產品開發）多踩一腳，我們都有如同任意門般的溝通管道以及循序漸進的計畫能夠輔助你達到你想要的職涯目標。 【實際接觸到的工作內容】 1.與數據科學家 (DS)、產品經理和其他技術團隊密切合作，設計並部署 AI 演算法與模型應用服務，解決 91APP 的各種商業問題 2.設計並開發高效且可擴展的模型服務，確保模型成果能夠在大規模生產環境中穩定運行，並可供產品線、第三方合作夥伴或品牌使用 3.針對不同商業場景與需求，設計與建置 AI Service，包含資料前處理、feature engineering 與 model inference 等，提供不同產品線、第三方合作夥伴或是品牌使用模型服務 4.學習 AI/ML 最新技術、進行實作，並透過實際商業問題的 PoC 評估可行性 5.實踐 MLOps 進行 Model Performance 監控、re-training 與模型版控 【期望你具備的基本條件】 1.至少 2 年 Data 或 1 年 Backend 相關開發經驗，具有 FastAPI 設計與開發經驗尤佳 2.熟練至少一項以下程式語言 e.g. Python, Java 或 Node.js 等 3.熟悉 DevOps 實踐，具備 git 版控、 CI/CD pipeline 使用經驗，確保模型和系統的穩定 部署與持續交付 4.熟悉 GCP，包含：Vertex AI、Bigquery、Cloud Functions、Cloud Run、GKE… 5.了解基本統計學、主流機器學習與 AI 演算法核心原理（Probability、LSH 、 similarity、Logistic Regression、KMeans、DNN、CNN、RNN、NLP…） 6.具備設計和實現安全、高效 API 的經驗，確保服務在多方合作的環境中保持高可用性和 數據安全性 7.保持對新技術的敏感度，熱衷於學習和分享最新的 AI/ML 技術與方法，並積極應用於解 決實際商業問題，以驅動創新和提升產品競爭力