About

Xingfan Xia is a seasoned software engineer with over 5 years of relevant experience in the field. Currently working as an Academic Technology Engineer at Carleton College, Xingfan has developed Python web crawlers to scrape course information from Carleton’s learning management system using Selenium and bs4. Additionally, he implemented a rendering engine to compile collected data as PDF documents to provide convenient access for faculty with PhantomJS and gecko driver. Prior to his current role, Xingfan worked as a Machine Learning Engineer at Apple where he contributed to on-device user profiling framework and built listener services to collect feedback and loggings to better track engagement of suggestions and prediction provided. He improved the current location model using more diverse data sources with better filtering and data sanitation and overhauled the topic prediction model with more up-to-date data sources and sanitized features using BERT language embedding models. Xingfan also developed an automatic machine learning pipeline which retrains models with data source updates, launches experiments A/B testing the models, and deploys if a specified improvement is observed. He is currently working on action prediction features which predict what a user intends to do at a particular time and location based on various signals and models. Xingfan has also worked as a Software Engineer Internship at Vevo where he built a Q&A chatbot using AWS Lambda, Python, and AWS Lex. He deployed the chatbot on Facebook messenger and Slack platforms. He also built a multi-seed recommendation engine with a data scientist on the team, deployed as an API service and a Web App connected with the existing user service system. Xingfan deployed a Jupyterhub system on an AWS P2 instance to facilitate the team's research and exploratory works related to machine learning and deep learning, typically made model training 6 times faster. He analyzed visuals of Vevo's music videos, generated playlists that are based on similar visual elements by topic modeling using machine learning. Xingfan holds a Bachelor's degree in Computer Science from Carleton College. His tech stack includes Software Engineer, ML, AWS, Amazon Athena, BERT, test, iOS, Scikit, Redis, Java, Pandas, Kafka, and Apache Hive. Xingfan is passionate about making changes and making stuff and is always looking for new challenges to tackle.