About

Xuancheng Fan is a highly skilled Machine Learning Engineer with over 9 years of experience in the field. He is currently working with IBM as a Machine Learning Engineer (Nature Language Understanding) where he is responsible for implementing and deploying new emotion/tone models using deep learning techniques such as CNN and LSTM. Xuancheng has also re-designed and refactored current NLP workflows’ logic flow to optimize its performance and quality, for example, aggregate document and sentence sentiment workflow. He supports cloud/dev engineers to fix all model’s compatibility and deployment issues. Prior to IBM, Xuancheng worked with Ericsson as a Data Scientist II where he led the improvement of performance of Tickets Solution Recommendation to 60% Top5 accuracy increase (20% to 80%) and 50% Top1 increase (20% to 70%). He also refactored and built a common asset for ticket solution recommendation which was selected to be one part of Ericsson Network intelligence, Ericsson’s AI solution. Xuancheng also led the Hardware False Fault detection system project, including project scope management, data preparation, feature engineering, and modeling. He designed the end-to-end solutions including modeling, data pipeline, life cycle management, and deployment and implemented the solution using Docker and pipeline tools (Kubeflow, MLflow). Xuancheng has also worked with Huawei as a Data Engineer (via Vertisystem) where he doubled the model accuracy and efficiency in error KPI recommendation system project after implementing data wrangling/engineering (e.g., Pandas, Numpy), data analysis to identify outliers from KPIs (time-series data) and utilizing machine learning models (e.g., KDE, BSCAN). This project led to a-million-dollar saving for the maintenance department. He also implemented a full range of data/programming tasks according to the requirement of projects, including data wrangling (e.g., filter outliers), exploratory data analysis (e.g., find out the relationship between features or cases), feature engineering, data science, and machine learning (e.g., provide AI solutions for root cause analysis or production line monitoring). Xuancheng also implemented testing, reporting, visualization, and big data service, e.g., Hadoop, Hive. Xuancheng holds a Master's degree in Management Information Systems, General from the University of California, Berkeley, and a Bachelor of Science in Information Security from Shanghai Jiao Tong University. He has expertise in ML, Deep Learning, CNN