• Processed over 10 GB data per day and generated regular reports with Airflow ETL pipelines.• Designed an internal library to collect, process and analyze data. Over 50% projects depend on it. • Collaborated with backend engineers and made the ETL pipelines of examination data from over 40 labs. • Maintained an existing website and added new features to explore more data insights for internals.• Trained a sequence labeling model with PyTorch to recognize Chinese food names on free text.• Built a model to detect users are likely to leave our service and sent retention messages automatically.

• Built a document search platform to store, analyze, and reuse knowledge about industrial manufacturing.• Led an intern to enhance search algorithm for semantic matching on our Solr platform.• Developed RESTful API with flask to detect keyphrases and NEs for all documents automatically.• Designed Text Annotation Support Service (TASS) with architects for human and machine labeling.• Implemented DAO to bridge TASS and DB by using MyBatis and cooperated with backend team smoothly.

• Implemented a greedy algorithm for labeling sentences with sense sequences.• Filtered the sentences containing NEs from Wikipedia; labeled NEs and sense information.• Visualized labeled sequences; utilized Unit Testing to verify the quality of the visualization tool.• Participated in several teams simultaneously; collaborated with different members effectively.