柯郁任 Yuren Ke

[email protected]

0916-990481

Taipei, Taiwan

Summary

  • Software engineer with 8+ years' experiences in mobile protocol stack, middleware and apps development in C.
  • ML engineer with 3+ years' experiences in NLP, Computer Vision, recommendation system.
  • Passionate about machine learning, data mining, NLP, Computer Vision and big data analytics.

Skills


Programming Languages

  • C/C++
  • Python
  • SQL
  • R
  • Swift
  • Java


Data Engineering / Machine Learning Tools

  • Scikit-Learn, Numpy, Pandas
  • Tensorflow, Keras, Pytorch
  • NLP: Gensim, Jieba, BERT
  • Visualization: pyplot, ggplot2
  • Similarity Search: Faiss
  • Workflow: Airflow
  • Cloud: GCP BigQuery, GAE, GCE, GCS, Dataproc, Dataflow, Data Studio
  • DB: MySQL, PostgreSQL
  • Search Engine: ElasticSearch
  • Distributed Computing: pyspark
  • Docker


Domain Knowledge

  • Machine Learning: SVM, Logistic Regression, Random Forest, GBM,...etc.
  • Deep Learning: CNN, RNN, LSTM, Transformer, BERT, Word Embedding, Object Detection, Reinforcement Learning, ...etc.
  • Mobile Protocols: GPRS/WCDMA/HSPA, TCP/IP

Education

National Chiao Tung University, Taiwan,Sep. 2003 - Oct. 2005

M.S. in Computer Science and Engineering

National Chiao Tung University, Taiwan,Sep. 1999 - Jun. 2003

B.S. in Computer Science and Engineering

Experiences


ML Engineer,May. 2022 – Present

RevoVision, Taiwan

  • Spam and profanity filter for online casino chat rooms
    • compared with the original design, the mistaken deletion rate is reduced by 20% ~ 30%
  • Automatic user nickname change audit system
    • accuracy: about 98%
  • Dealer Photo Cropping Tool
    • automatically generate thumbnails that can be used as game room dealer headshots
    • automatically horizontally flip the thumbnails to make each dealer face the same direction in the photo
    • save manpower and time required for manual cropping
  • Assist in risk analysis of various online casino games

Senior Algorithm Engineer,Jun. 2018 – Aug. 2020

PIXNET, Taiwan

  • Adopt BERT to Implement auto reply feature for customer service department
  • Implemented search by Image service
    • one can search similar images and related BLOG articles in PIXNET with a photo
  • Word suggestion API for PIXNET bloggers
    • will provide recent hot keywords or names of POIs for bloggers to add hashtags  according to what they type
  • Algorithms implementation to recommend hot related articles
    • 4%~8% CTR improvement
  • Voice check-in system for PIXNET HACKATHON 2018
  • Photo Style Transfer for PIXNET Social Summit 2019

Postgraduate,Jan. 2018 – Apr. 2018

Taiwan AI Academy, Taiwan

  • This bootcamp covers traditional ML and Deep Learning technologies. 
  • Programming Language: Python / R
  • Final Project: Photo Teller - Image Caption
    • http://www.phototeller.ml/
    • Merit Award of Final Project Presentation of Taiwan AI Academy
    • Adopt CNN to extract image features, and attention model to help LSTM language model generate captions
    • Implement in Tensorflow(main model) and Keras(image feature extraction)
  • Competition: T-brain file agnostic malware detection
    • https://tbrain.trendmicro.com.tw/Competitions/Details/1
    • rank 20% in competition, AUC 0.924134
    • Adopt XGBoost with GridSearch to perform logistic regression and predict if a file is a malware

Principal Software Engineer,Sep. 2012 – Mar. 2015

MediaTek, Taiwan

  • Responsible for 3G L2 Network Protocols software maintenance
  • Implemented L2 RLC/BMC part for dual SIM/dual standby projects

Senior Software Engineer,May. 2007 - Sep. 2012

MStar Semiconductor, Taiwan

  • Responsible for 2.5G L2/L3 Data Plane Network Protocols
  • Designed and modified the MStar proprietary TCP/IP and DCHP to support Wi-Fi feature
  • Developed new data delivery mechanism to enhance data throughput of Wi-Fi. Throughput raised by 50% and MIPS consumption reduced by 40%
  • Supported the multimedia team for the Tudou HTTP/RTSP video streaming app (www.mstarsemi.com/tw/news.php#64)

Software Engineer,Feb. 2006 - Apr. 2007

BenQ, Taiwan

  • Ported a 3rd party QR code scanner solution and implemented the corresponding app
  • Ported a picture rendering solution (Scalado) to speed up the pictures / thumbnails rendering

Courses & Certificates

  • Coursera    Data Engineering, Big Data, and Machine Learning on GCP                                                                 Sep. 2021
  • Coursera    Smart Analytics, Machine Learning, and AI on GCP                                                                                Sep. 2021
  • Coursera    Building Resilient Streaming Analytics Systems on GCP                                                                        Aug. 2021
  • Coursera    Building Batch Data Pipelines on GCP                                                                                                         Jul. 2021
  • Coursera    Modernizing Data Lakes and Data Warehouses with GCP                                                                       Jul. 2021
  • Coursera    Leveraging Unstructured Data with Cloud Dataproc on Google Cloud Platform                                 Jul. 2018
  • Coursera    Google Cloud Platform Big Data and Machine Learning Fundamentals                                              Jun. 2018
  • Taiwan AI Academy    courses                                                                                                                     Jan. 2018 - Apr. 2018 
  • Udemy        R Programming A-Z: R for Data Science with Real Exercises                                                                  Jan. 2018
  • Udemy        The Complete IOS 10 & Swift 3 Developer Course                                                                                  Oct. 2017
  • NTU@Coursera    Machine Learning Foundations & Techniques                                                                              Jun. 2017