Cole Springer

University of Illinois Springfield | M.S. Data Analytics, GPA: 4.0 Expected 12/2026

Relevant Coursework: Introduction to Machine Learning, Deep Learning, Big Data Analytics, Statistical Computation, Advanced Database Concepts, and NoSQL Databases.

University of Southern Mississippi | B.S. Mathematics, GPA: 3.7 05/2021

SGS North America | Columbus IN, USA 08/2022 – 12/2023

Software Engineer (Full Time, 40 hrs/week)

Developed a Java REST API to extract information from customer application queries and inject relevant processed inputs from an internal calculation engine.
Designed and deployed a customer-facing database application to ensure federal regulatory compliance regarding the consistency of carbon and energy content in fuel samples.
Migrated legacy data transmission protocols (CORBA) to gRPC, significantly improving the scalability and maintainability of API connections.
Updated and maintained SQL tables to satisfy evolving customer reporting requirements and ensure data integrity.
Created unit tests to verify code reliability.

Brain Tumor MRI Classification | University of Illinois Springfield 05/2026

Built a hybrid Convolutional Neural Network and Vision Transformer (CNN-ViT) model in PyTorch for a four class brain tumor MRI classification using the Kaggle Brain Tumor MRI dataset.
Created an image preprocessing and evaluation pipeline with grayscale conversion, normalization, train/validation/test splitting, Optuna hyperparameter tuning, and baseline comparisons against CNN and ViT models.
Achieved 93% test accuracy and 93% F1 score, outperforming CNN and ViT baselines while generating class specific evidence maps to evaluate model interpretability and identify limitations in explanation reliability.

Fake Job Posting Detection | University of Illinois Springfield 12/2025

Developed a multi-input Neural Network model using Keras and Word2Vec embeddings to detect fraudulent job listings within the EMSCAD dataset.
Implemented text vectorization for multiple description fields and utilized class weighting to successfully handle severe dataset imbalance (5% fraud rate).
Achieved 86% Precision and 76% Recall, significantly outperforming a Naive Bayes baseline (24% Precision) and validating the model's viability for sensitive fraud detection tasks.

Languages: R, Python, Java, SQL (Oracle, SQLite), Bash.
Data Science: Deep Learning, Machine Learning, Computer Vision, CNNs, Vision Transformers, NLP, Apache Spark, Apache Hadoop, Statistical Analysis, Data Cleaning.
Frameworks & Tools: Docker, Linux, Tidyverse, NumPy, Pandas, Keras, PyTorch, Optuna, Spring Boot, Hibernate, gRPC, React, Git, Azure DevOps.

Education