I am an AI and machine learning practitioner and an M.S. candidate in Psychology: Data Science in Human Behavior at the University of Wisconsin–Madison. My work combines applied machine learning, document intelligence, behavioral data science, and experimental research.
I build end-to-end AI systems, including multi-agent document-processing workflows, transformer fine-tuning pipelines, LLM evaluation harnesses, computer vision classifiers, and predictive models. I am particularly interested in reliable, measurable, and cost-efficient AI systems.
My master’s capstone with American Family Insurance focuses on improving document intelligence workflows and reducing AI inference costs. I have designed reproducible LangGraph pipelines for document routing, entity extraction, classification, prompt experimentation, deterministic scoring, dataset management, and deployment.
I have practical experience with Python-based ML development using PyTorch, Hugging Face, scikit-learn, ONNX, FastAPI, Docker, and cloud tooling. My work includes model calibration, cross-validation, error analysis, SHAP feature importance, OCR-based classification, and evaluation of LLMs against traditional ML baselines.
My research background includes cognitive neuroscience, behavioral prediction, psychometrics, and EEG data analysis. I have contributed to research on chatbot truth and deception, cognitive workload, sleep quality, and communication apprehension, and I presented findings at the 2025 Midwestern Psychological Association conference.
I also bring five years of logistics and systems operations experience, with strengths in quality assurance, operational documentation, root-cause analysis, and iterative process improvement. I am open to relocation and seek opportunities where I can apply data science and AI engineering to meaningful real-world problems.
