Projects | Jeremiah Dy | Professional Portfolio

AVAA (Vaccination Helper Webapp) 2025-08-02

AVAA is a straightforward full-stack app designed to simulate a website assistant that helps patients and medical staff schedule and manage vaccination appointments. It also provides simple data visualizations to report on appointment trends and patient demographics.

JavaScript Fullstack Development Node MongoDB Express React Redux Bootstrap

Hotel Registry with AI-powered Chatbot 2025-07-04

This application is designed to simulate a hotel booking website that spans a multi-hotel network. It includes a built-in chatbot that understands and responds to user queries in natural language, powered by OpenAI's advanced NLP models.

Java JavaScript Fullstack Development Spring-Boot PostgreSQL OpenAI React JWT (JSON web token) React

IT Support Ticket System 2025-06-06

This multi-service application simulates a company's internal IT support ticketing system. It allows regular employees to submit IT help tickets, which are reviewed and either approved or rejected by managerial staff, resolved by IT support staff, and closed or reopened by the original submitter. This system is built using a micro services architecture, with each service handling a distinct area of functionality.

Java JavaScript Fullstack Development Spring-Boot MVC Application (Model-View-Controller) MySQL ActiveMQ

Machine Learning Approaches to Breast Cancer Classification with All of Us Data 2024-05-10

The All of Us program is a program organized by the (United States) National Institute of Health which serves to aggregate, anonymize, and make available patient health data for research projects. Using the All of Us database, I built several classification models to predict malignancy in breast cancer patients, as well as compare differences in model performance between two differentiated dataset groups.

Python Machine Learning All of Us Program MLP SVC Random Forest Adaboost Gradient boosting models

Receipts Parser using ChatGPT 4.0 2023-12-11

Working with three other of my ICS 438 classmates, we constructed a program which would parse the raw text obtained from using optical character recognition (OCR) on random consumer receipts into JSON objects using OpenAI's large language model, ChatGPT. It would then provide analytics on the JSON data and use it to train a KNN model from the FAISS library (open-sourced from Facebook AI) to classify new receipts into categories. The data from analysis would be used to create a simple visualization using Streamlit.

OpenAI ChatGPT Python Pydantic JSON Machine Learning Github Optical Character Recognition FAISS

Titanic Survivability Calculator 2022-12-14

With three of my other classmates in ICS 484, we built a website that would calculate an individual's survival rate had they been on the Titanic when it sunk (April 1912).

HTML JavaScript R Website Design Data Science Data Visualization LPM