Graduated from Boston University (MS in Electrical and Computer Engineering). Currently working as a Software Engineer in Pegatron Corporation based in Taipei, Taiwan.
New Taipei City, Taiwan [email protected] 0909820803 ChihWeiTung tungchihwei
Python JavaScript Machine Learning Natural Language Processing Tensorflow RESTful API ElasticSearch ReactJS Docker MySQL MongoDB Kafka Jenkins Airflow pySpark Django(Python) FastAPI Flask api SparkSQL
November 2022 - Present
1. IoT data transmit structure
・Transmit IoT data by using Kafka and MQTT (Python).
・IoT data analysis.
2. Customized dashboard platform development (React, Python)
Backend (Python)
・MongoDB table schema design for dashboard data.
・Restful api for save/edit/delete data in MongoDB.
- Dashboard permission modification (from MongoDB).
- Get user list (Django SQL).
- Dashboard config saving api (from MongoDB).
- Dashboard list filtered by the current login user api (from MongoDB).
- Dashboard delete api (from MongoDB).
Frontend (React)
・Provide panel that can adjust size and position of charts.
・Provide various charts and display setting for each chart.
・Provide data transform function. Data can be adjusted according to the needs (group by, merge, ...).
・Provide static/streaming data transmission modes for user to transmit data through api or websocket.
3. Server performance monitor (Flask api, Linux)
・Total server/container cpu usage rate.
・Total server/container memory usage rate.
・Server io usage api.
・Each container cpu usage rate and memory used.
June 2021 - October 2022
・Create ETL for cleaning daily customer transaction data by using Spark SQL.
・Build and optimize classification model for predicting possible customer repurchase dates.
・Transaction data analysis and visualization.
・Video analysis by using Google Video AI and Azure Video Analyzer.
・Provide text to speech function by using Azure Text to Speech.
・Fine-tune NER model for optimizing the result of text to speech.
September 2019 - March 2021
1. Social Media Analytics
・Generate and visualize daily/weekly hot topic by analyzing news from various sources.
・Crawl, clean and combine similar news from different sources.
・Achieve 90% f1-score by building model for classifying news to 12 classes (BERT+LSTM).
・Extract keyword and summary for each news, and the results are used for daily/weekly hot topic extraction.
・Store data in Elasticsearch, and create RESTful API for visualizing results on our website.
2. Chatbot Platform
・Analyze customer service chatbot dialogue and performance by using BERT and SVM.
June 2019 - August 2019
・Teach Artificial Intelligence and Machine Learning.
・Skills students got after these courses are Linux, Python, TensorFlow, Keras and so on.
November 2015 - July 2016
・Facilitate Proactive Database Project execution co-held by YZU and III.
・In charge of the deployment of Toronto, a MongoDB-based big data processing system developed by III, in YZU’s Data Center.
・Web crawling and data analyzing for YZU by using Toronto.
・Due to the demand for a big data analytics platform in YZU, conducted data analysis by using Toronto through Restful API. And test the system for potential errors.
2017 - 2019
2011 - 2015