Turning complex data into intelligent systems.
I'm Michael, a data science student at Northeastern building ML-powered products, from a published iOS app to real-time analytics platforms serving hundreds of users.
I work where data, machine learning, and software meet, turning messy information into tools people actually use. I'm especially drawn to applied AI: embeddings, retrieval systems, and the quiet infrastructure that makes intelligent products feel effortless.
Things I've built
Real-time biomechanical feedback platform that replaces manual video analysis for track & field coaches. Shoe-mounted sensors stream ground-contact time, fly time, step counts, and hurdle splits to a live analytics dashboard.
Guitar ear-training app with daily chord-recognition puzzles, progressive difficulty, and competitive multiplayer. Built natively with Swift and SwiftUI.
Deep RL agent that learns to play Quoridor through behavioral cloning and self-play. A PPO policy over a CNN board encoder reached a 97% win rate against a random bot and consistently beat a heuristic opponent.
ML-powered student-experience matching using semantic embeddings to connect students with personalized opportunities. Serving 500+ daily users at 99% uptime.
Fifty-year study of how software-engineering research questions shift over time, modeling ~11,000 ICSE papers with BERTopic to trace topic drift and validating detected onsets against dated field milestones.
Detects and classifies thirteen types of logical fallacies in text, comparing TF-IDF, deep averaging networks, and fine-tuned transformers. RoBERTa reached the strongest macro-F1.
Document-analysis framework that extracts insights from business filings using NLP and pattern recognition across unstructured data.
Designed and built the consulting agency's site with Astro and Sanity CMS, featuring leadership profiles, project portfolios, and automated deployment via Vercel.
Retrieval-augmented generation tool that turns personal notes into an interactive knowledge base, tuned for accuracy through systematic embedding-model testing.
Where I've worked
Data Chief
Generate Product Development Studio
As Data Chief, leading four developers in building StrideTrack, a real-time biomechanical analytics platform for track & field coaches. Driving architecture through code reviews and pair programming in FastAPI and React, and defining API endpoints and database schemas from client requirements.
Teaching Assistant
TA for Database Design (CS3200), supporting 200+ students with SQL, data modeling, normalization, and query optimization through office hours and grading.
Software Engineer
NExT ConsultingLed development of ML-powered platforms and manufacturing data-flow systems, architecting scalable REST APIs with Django and PostgreSQL and real-time production dashboards.
Student Innovation Developer
Built a licensure-tracking system for the Office of the Provost, monitoring regulatory requirements across 51 programs in 63 U.S. territories via automated scraping and AI-powered analysis.