Data Science • Machine Learning • Applied AI
I am a machine learning researcher and data scientist focused on applied AI systems, statistical learning, and model evaluation. My work spans certified anomaly detection, concentration inequalities, regression modeling, retrieval-augmented AI assistants, and analytics systems that turn complex data into reliable decisions.
Research
Submitted
A certified anomaly detection framework for IoT network security using concentration inequalities for weighted Poisson processes, designed to provide formal false-positive guarantees while supporting real-time inference.
Submitted under review
A negative-binomial concentration framework for overdispersed IoT count traffic, extending certified anomaly detection beyond equidispersion assumptions commonly used in Poisson-based models.
Ongoing
A data-driven framework for collaborative resource sharing across institutions, exploring how algorithmic matching can reduce backlogs while preserving fairness, quality standards, and resource efficiency.
Featured Projects
Applied AI
A deployed RAG assistant that lets users ask natural-language questions about Eddie's research, projects, skills, and portfolio background.
RAG • LLMs • Azure AI • FastAPI • Retrieval Systems
Algorithm Design
A capacity-constrained matching framework for pairing institutional needs with contributor capabilities, developed through a CRL cooperative cataloging use case
Algorithms • Resource Allocation • Matching Logic • data-anonymization
Data Olympiad
A 1st-place energy forecasting project using XGBoost, SHAP explainability, residual diagnostics, and savings simulations across a large commercial building portfolio.
Forecasting • Energy Analytics • Machine Learning • Business Impact
Healthcare ML
A healthcare machine learning case study comparing supervised classifiers, neural networks, ensemble models, and exploratory anomaly detection for benign-versus-malignant tumor prediction.
Classification • Model Evaluation • Healthcare Data • Supervised ML
Anomaly Detection
An unsupervised learning project that combines traffic-pattern clustering and probabilistic anomaly detection to identify operational patterns and unusual energy-grid behavior in smart city systems.
Clustering • Anomaly Detection • Smart Cities • Unsupervised ML
Business Analytics
A predictive analytics case study for identifying likely purchase sessions using browsing behavior, visitor context, threshold tuning, and model interpretation.
Gradient Boosting • Conversion Analytics • Threshold Tuning • Feature Importance
Regression Modeling
Built a neural-network regression workflow for house price prediction, benchmarked it against traditional models, and analyzed prediction errors in real dollar terms.
Neural Networks • Regression Benchmarking • Error Diagnostics • Feature Importance
Visual Analytics
Interactive Tableau dashboards and Power BI visual analytics work for airline operations, workforce survey analysis, location intelligence, and business reporting.
Tableau • Power BI • Dashboards • Data Storytelling • Business Intelligence
Skills
My technical toolkit spans data preparation, machine learning, model evaluation, AI engineering, business intelligence, and research computing, with a focus on building reliable systems that turn data into practical decisions.
Certifications
Microsoft Certified
Cloud Foundations
Foundational credential covering Azure cloud concepts, services, governance, pricing, and core cloud architecture.
Verify credentialMicrosoft Certified
Analytics & BI
Professional credential focused on data modeling, visualization, dashboarding, analytics, and business intelligence reporting.
Verify credentialMicrosoft Certified
MLOps & AI Engineering
Associate-level credential aligned with deploying, monitoring, and managing machine learning systems in production-oriented workflows.
Verify credentialPresentations
Presented at Canisius University, highlighting AI-assisted approaches for addressing cataloging backlogs and improving information access workflows.
Old Main OM 102 • Center for Research Libraries Global Resources Network
Presented with Prof. Sana Spektor at Canisius University, connecting probability theory, concentration inequalities, and machine learning for cybersecurity anomaly detection.
Science Hall • Canisius University
Achievements & Recognition
Competition Award
Data Science Olympiad • April 2025
Awarded first place for a competitive data science project involving analytical problem-solving, predictive modeling, and communication of findings.
Research Fellowship
2025 • Canisius University
Fellowship support connected to research on concentration inequality-based machine learning methods for certified anomaly detection.
Academic Recognition
Academic distinction
Recognition for academic excellence and sustained academic performance, included as part of the portfolio knowledge base.
Contact
I am open to data science, applied AI, machine learning, and research-oriented opportunities where analytical thinking, model development, and clear communication matter.