Chengdu, Sichuan, China
Personal website: xukun-liu.com Great passion with data and dedicated to business analytics Seeking to leverage data analytical skills to explore business insights and improve operational performance as a Data Analyst/Data Scientist • Proficient in R, Python, SQL, Tableau, • Excel, VBA, HTML • Jupyter Notebook, RStudio, Django, Hadoop, Spark • Time series analysis, Linear/logistic Regression, A/B Testing, KNN, Random Forest, CNN, NLP
- Business Intelligence Competency Center, aiming to use information technology to speed up the company's R&D (Research and Development) process and empower the management team, providing optimization suggestions for business planning and resource allocation. I am serving as a data analyst, providing technical support to teams within the department for statistical analysis, mathematical modeling, data visualization, etc. - Participated in the design and implementation of the company's internal project monitoring system. Integrated data visualization tools to track project progress and manage 300+ R&D projects in 10+ countries around the world - Designed interactive data dashboards (Python) with colleagues from France and US to empower the management team to analyze projects from different dimensions such as regions, functionality, departments, roles, etc. - Identified abnormal data and metrics through programs (Python, SQL), provided early warning to the operation teams, and advised local manager on intervening project plan deviations promptly Optimized business data reporting and analysis process - Automated weekly project data gathering, processing, aggregation, and archiving using Python, Airflow, and other tools - Responsible for daily operation and monitoring of data platforms (SQL), effectively reduced data aggregation lead time by 30%, improved data quality, and significantly reduced manual errors - Created new tools to solve our stakeholders’ pain-point in daily work: by utilizing Elasticsearch, developed an API, and allowed the R&D team to efficiently search and retrieve records from their archived documents
• Cooperated with stakeholders, identified business and technical requirements for credit card transaction testing tools • Fixed bugs and rewrote core functions of test automation tool’s code in Python, reduced test execution effort and time by 50% • Built an app with user interface (Tkinter) for API test configuration, decreased preparation time by 70% and avoided human errors • Collaborated with development team to locate errors, used Postman to investigate API issues in 200+ transaction testing cases
• Developed a web-based GUI (Django) to query database with one million records, reduced query/analysis cycle time by 30% • Designed three interactive visualization dashboards with Bokeh/Tableau, enabled users to create plots based on customized needs • Constructed a fully connected neural network model for crime-related social media content detection, gained 80% accuracy rate • Applied and trained different NLP models to classify crime-related text by offense types, achieved 65% accuracy rate