Portfolio
Data science and machine learning
I'm an AI engineer and analytics translator: I turn messy business and education problems into clear questions, models and tools people actually use. By day that's enterprise analytics at Sysco; on the side, AI coaching pipelines at Cignition.
- LLM pipelines and RAG with LangChain, sentence-transformers and ChromaDB
- Deep learning and transfer learning in PyTorch and TensorFlow
- Regression, classification and ensembles with Scikit-learn and XGBoost
- Python, SQL, Pandas, Databricks and Tableau for analysis and reporting
- Explaining results clearly to technical and non-technical audiences
MS in Data Science, Eastern University (expected 2027) · Data Science & ML Engineering, 4Geeks Academy (2025) · Advanced ML and Agentic AI, Interview Kickstart (2026)
RAG tutor coaching pipeline
Retrieval-augmented AI that grounds coaching feedback for tutors in the company's own instructional guidance. Built for Cignition; internal tool, so no public link.
Skin lesion classifier
ResNet50 transfer learning on 25,000+ dermoscopy images across 8 diagnoses. Handled a 54:1 class imbalance and lifted macro F1 from 0.44 to 0.67.
Salary predictor
How education, skills and early experience shape starting salary. Full lifecycle from cleaning to a deployed Flask app.
FitCast
Predicts calories burned or suggests workout length from biometric data, in an end-to-end Scikit-learn and TensorFlow pipeline.
Podcast listening time (Kaggle)
XGBoost model predicting minutes listened, built for a Kaggle competition.
Real estate data visualization
Cleaning a messy housing dataset and visualizing what drives price.
Image classification: cats and dogs
A convolutional neural network for image classification, with learning-rate and batch-size experiments.
