Daniel Kelly

Driven and detail-oriented Computer Science student with a passion for software development and problem-solving. Proficient in multiple programming languages including Java, HTML and CSS with hands-on experience in developing web applications and implementing algorithms. Adept at collaborating in team environments, demonstrated through successful project contributions while leveraging strong communication skills to deliver effective solutions. Seeking an internship to apply theoretical knowledge in a practical setting, further enhance skills, and contribute to innovative tech projects.

Brokerage Account (Java)

• A simple implementation of a brokerage account that includes various classes and methods to resemble a real world investment account.
• Includes Stock, Cash, Date, and Customer classes.

Stock Market Stock Trader (Java)

• Using type hierarchy and various data structures to simulate a trader that can buy and sell stocks
• Simulate a market that maintains all buy and sell orders for a stock as well as the prices for those stocks

Data Structure Implementations (Java)

• Arrays, Stack/Queue modifications, BST, Hash Table(separate chaining), Sorting Algorithms, Undirected Unweighted/Directed Weighted Graphs

AI Stock Predictor (Python, PyTorch)

• Using a company's social media trends, we used machine learning models to try to predict the volatility of their respective stock. Scraped and cleaned data using Apify and Python.
• Used PyTorch to train models for Tesla, Nintendo, and GoPro based on user engagement with their social media posts (Instagram, Twitter, and TikTok).
• Utilized Parallel Processing (Apache Spark) to scrape and clean media data to look for percentage changes in stock price and trading volumes while assigning binary indicators to above average changes.

Booking Form (HTML/Bootstrap)

• Using Bootstrap framework to implement a simple booking form that include name, email, date of appointment, etc

To Do Application (HTML/CSS/Javascript/Vue)

• Simple to-do web application that utilizes Vue framework to create to-do tasks that can be completed, edited, completed, etc
• Deployed using VScode IDE, GitHub, and AWS

AI Comic Strip Generator (Python, machine learning models)

• Scrape all xkcd comic strips using a parallelized python script. Use Globus to transfer all necessary files to and from CWRU's HPC cluster.
• Used Stable Diffusion XL (open source machine learning model) and VQGAN-CLIP to generate images from input text. Results found GPU utilization was necessary for quality image generation