- Comparison of Two Question Answering Pre-trained Models (2024): Researched the Transformers architecture and compared two identical QA models using F1-score and Exact Match metrics. Used Hugging Face Pipelines to load the pre-trained models, formatted the model outcomes, tracked time during the process, and used Matplotlib for visualization. Python, NLP, Pipeline
- Jason Scott Shoes Full-stack E-commerce website (2024): Achieved a functional website by connecting with two different frameworks and languages via REST APIs. Applied JWT to authenticate users. Implemented sync function to store the data for registered users while the visitors’ data is stored in cookies. Django, React, Postgres, JWT, Cloudinary
- Spring To Do Task (2024): Applied different dataset configurations, JDBC, and JPA. The app is structured into different sets of layers: business, data, and presentation layers. Used Docker to run database service and the Server-side engine to render the app. Spring, Docker, JDBC, JPA, Java, Thymeleaf, Bootstrap