Implemented a serverless ETL pipeline using AWS S3, Glue and Lambda to convert large-scale XLSX datasets into Parquet format, achieving 218% storage savings and 156% faster query performances.  

Developed a system integrating OpenAI GPT and AWS Redshift Spectrum to translate natural language queries into SQL for real-time data retrieval and analysis. 

Created an interactive US county loan analysis map using Tableau with multiple filters, and embedded the visualization into the company website for real-time insights.

Led an A/B test initiative to optimize editor engine for our product, including editors comparison and user data analysis. Executed data processing using Hadoop MapReduce in Java, and utilized PowerBI for visualizing results on performance, user experience, and compatibility testing, leading to a 10% increase in customer growth.

Developed a real-time Risk Alertness-Forecasting Model using Probability & Impact Matrix and GMM, integrated with Apache Airflow and Databricks to automate daily risk alerts. Incorporated manual review and additional features, continuously retrained the Random Forest model, reducing task delivery delays by 60%.

Utilized SVM algorithm with MobileNet2 in Python to rebuild the CCC code recognition model. Collected and preprocessed a dataset of 0-9 numbers for training and testing, increasing the accuracy from 96% to 99%.  

Trained the YOLOv5 model using Spark on AWS EMR and implemented it in a QR code scanning program for forklifts to digitize materials and streamline management.

Used ResNet50 and BEiT for painting style classification to enhance the Art Appreciation Algorithm.  

Trained LLAVA with the style classification model and SemArt dataset, integrating RAG for knowledge retrieval, boosting accuracy by 360% (Rouge-1) and 53% (Meteor) over direct image appreciation.

Architected the backend using Spring Boot and expertly crafted the frontend with React, Redux, Node.js, developed real-time conversation, user profile and transaction posting pages.

Executed cloud-based DevOps: Deployed the application through a Kubernetes cluster on Azure, oversaw the storage and management of Docker container images within the AWS environment.

Degree: M.S. in Computer Science specialized in Data Science and Machine Learning

GPA: 3.8/4.0

Courses: Data Visualization, Tools and Models for Data Science, Deep Learning for Vision and Language, Informational Retrieval

Degree: B.S. in Computer Science Honor Program (with First Class Honor)

Courses: File and Database Management, Visual Recogition, Intelligent Systems, Numerical Methods, Machine Learning, Reinforcement Learning, Statistics

Programming: Python(NumPy, Sklearn, Panda, PyTorch, TensorFlow), R, JavaScript, C

Big Data: MongoDB, DynamoDB, Solr, MS Excel