Marvin Hodges

Aspiring Software Engineer intern with expertise in machine learning and front-end development, pursuing a Bachelor's in Computer Science with an AI concentration at Wilmington University. Proficient in PyTorch, scikit-learn, pandas, and neural network architectures, with hands-on experience building and deploying AI solutions like CNN-based classifiers and Retrieval-Augmented Generation chatbots integrated with React front-ends. Developed strong leadership, communication, and time management skills through full-time retail management while excelling academically. Eager to apply technical and collaborative skills to innovate in AI-driven systems and user-focused software solutions.

• Retail Chatbot, Developed an AI chatbot for retail associates that utilizes a Langchain/Retrieval-Augmented Generation (RAG) pipeline. The bot efficiently retrieves information from a vector database of company process guides to provide immediate and helpful responses, improving associate performance and knowledge access.
• IDMYBIRD, Developed a Convolutional Neural Network using PyTorch, trained to identify over 200 different species of birds with over 90% accuracy. The model is deployed via a ReactJS frontend, which calls an API to return the prediction and update the UI.
• Tumor Classification, Utilized a patient dataset to build a data science project in Jupyter Notebooks, employing various machine learning models (Linear Regression, XGBoost, and Logistic Regression) from the scikit-learn library to classify tumors as malignant or benign. The models were trained and evaluated using data points such as tumor area, compactness, and perimeter.