Founded a Google Developer Group at St. John's University. Lead and organize technical events.

Selected to work with IBM engineers on active development of the OSCAL Compass compliance automation toolkit. Implemented bug fixes, schema improvements, and documentation updates across multiple GitHub issues

Lead full-stack development of a clinician dashboard with React, Node.js, GraphQL, and SQL, streamlining ICD-10/CPT coding and review workflows. Integrate machine learning methods to deliver accurate medical code predictions and AI-generated clinical summaries, reducing manual workload for clinicians. Build scalable infrastructure with MySQL, Neo4j, and Milvus, ensuring reliable performance and seamless data retrieval.

Design and implement end-to-end NLP pipelines for 10K MIMIC-IV clinical notes, including preprocessing, tokenization, and evaluation workflows. Fine-tuned and evaluated BioBERT, achieving an F1 score of 0.84 for clinician burnout detection. Applied sentiment analysis, named entity recognition, and topic modeling to uncover linguistic patterns linked to psychological stress in ICU documentation.

Designed and trained a multi-classifier ML pipeline (LightGBM, XGBoost, Random Forest, MLP, SVM) to detect post-quantum cryptographic attacks in V2X networks, evaluated across 500K handshakes across six attack types

Achieved 0.968 AUC using cryptographic-layer telemetry features vs. 0.533 AUC for network-only baselines, demonstrating the value of feature domain selection in security ML.