Yujun Liu

A recent graduate of a master's program in data science and data engineering. Passionate about discovering trends in data and constructing workflows to transform raw data into knowledge.

Experience includes analyzing genomic data for a laboratory using statistical techniques in Python and R, and migrating pipelines from SQL Server to Azure using Python and SQL.

I am seeking a collaborative and fast-paced work environment where I can further develop my skills and make positive impacts for both the company and its clients.

8 months of experience at Bond Brand Loyalty, where tasks involved building pipelines in Microsoft SQL Server and migrating them to Azure Databricks.

• Collaborated with an agile team to meet biweekly deadlines
• Translated SQL code to Pyspark for over 20 tables, taking care to maintain the same transformation logic
• Assembled and tested pipelines on Azure data factory
• Proposed automation strategy to quicken production of JSON metadata files

1 year of experience at The Centre for Addiction and Mental Health, helping answer research questions by cleaning raw data and applying statistical tests to calculate p and r^2 values.

• Researched analyses in similar works to determine appropriate methods to analyze data
• Cleaned and normalized various data types for analysis, including multimodal and circular
• Managed time across multiple projects to ensure all deadlines were met
• Presented results by visualizing data in plots formed by ggplot2 and plotly in R

MSc Data Science and Engineering, University of Dundee, graduated 2023:

Relevant coursework and topics included:

• Gathering customer requirements, designing a star schema, then collecting and merging customer data from multiple sources in Python to build a data warehouse for the customer based on the designed schema
• Collaborated to present relational vs. non-relational databases, focusing on RavenDB and how it compares to alternatives
• Analyzed distributed data by using virtual machines on Google Cloud Platform, Docker, and Hadoop
• Set up microservices to construct a streaming service, using GitHub and Jenkins to automate builds

Master's Project was to improve upon a U-Net convolutional neural network model used for differentiating parts of bone images, to help forensic analysts estimate age

• Carefully annotated bone images, to ensure that the training data for the model was correct
• Implemented pre-trained architectures in Google Colab, using tensorflow and keras
• Researched appropriate performance metrics to monitor to assess model performance
• Analyzed effects of design changes on model performance to identify and fix weaknesses
• Presented improvements to the model through a written report and verbally at a seminar