Cake Talent Search

ESD EMI BCI/RS 微控制器 STM32 PIC Arduino 8051 Asm C 工具 Altium Designer OrCAD/Allegro KiC AutoCAD 工作經歷 Battery Systems Engineer Zitara Technologies 八月六月 2024 San Francisco, CA, the United States Taipei, Taiwan (Remote) ● 電池演算法展示板重新設計與製造： 在不改變設計架構與底層韌體規格要求下，實現 1) 改善電芯電壓輸入電流平衡，電流差異

At portto, we aim to make blockchain technology simple and accessible for everyone. DEXON Foundation, VP of Engineering, Mar 2018 ~ May 2019 DEXON is a native blocklattice technology designed to become the infrastructure for the future of the decentralized internet. At our core, DEXON is developing a consensus algorithm that is infinitely scalable, low-latency, and Byzantine fault tolerant. I lead the development of DEXON related products and planned its ecosystem. COBINHOOD, VP of Engineering, Sep 2017 ~ May 2019 Other than being a global top 20 cryptocurrency exchange, at COBINHOOD, we pride ourselves on our insight

the ISP pipeline and tuning parameters for various sensors. Reconstructed ISP algorithms using TensorFlow operators optimized for NPU compatibility, ensuring functional equivalence. Compiled high-level neural network models into low-level representations , profiling each operator for quantization errors and execution time on the NPU. Optimized the entire algorithm to ensure seamless deployment on the NPU. Automated the noise reduction parameter optimization using AI-driven techniques. Developing an AI algorithm with adjustable noise reduction intensity and successfully deployed the algorithm on NPU hardware. Improved the time complexity of Bayer noise reduction by 25x and enhanced

platform Utility of Marvell/Microsemi Ethernet-Switch chip and driver function Ethernet-Switch board verification and bring-up Development and Maintenance of QGD NAS-Switch bounded system solution Provide JSON RPC/Restful API for backend/frontend application Research Assistant • Institute of Information Science, Academia Sinica JanJun| Taipei, Taiwan Blockchain performance and efficiency testing platform base on Docker Swarm Classroom renting system of FinTech NTU organization by Ruby on rails Implementation of self-tuning PID controller algorithm on self-driving car Education National Taiwan University Bachelor's Degree, Department of Computer Science and Information Technology • SepFeb. 2019

Kao Chiang RD • NLP engineer，TW • [email protected] NTU ME #ML #NLP #Recommender Skills Software > ML: Tensorflow, Keras, Xgboost, LightGBM, Pytorch > NLP: Rasa, spaCy, Gensim, GluonNLP, GloVe > Web: Flask, Django, Vue, SQL > APP: Swift, Android Studio > Others: Docker, docker-compose, Spark Combination There are lots of open source tools and platforms to help us to develop the project, but how to combine each tool or platform well (efficiency & security) is the point. With some experience of organizing an whole project by myself, I am good at making good use of the newest technology

Focus on MCTS, optimization theory, reinforcement learning and statistical techniques. Optimization Programming • Build a tree-based reinforcement learning optimization algorithm applying ensemble dispatching rules and Monte Carlo Tree Search (MCTS) to solve scheduling problems with huge size under short time constraints /* Patent ID : TWI*/ • Develop a heuristic algorithm method(Lagrangian relaxation method) to solve scheduling problem in a large scale for searching a good solution in an effective way. • Develop a integrated system contains efficient meta-heuristics algorithm such as GA, PSO with Chaos-based initialization, ACO and AFSA(Artificial Fish-Swarm Algorithm) to solve classic