Goodreads RecSys & Market Segmentation
876K users / 223M rows on HDFS Dataproc (CSV → Parquet). MinHash LSH for top-100 book similarity pairs; ALS with Spark MLlib ranking metrics and implicit signals (is_read, is_reviewed).
Portfolio
Models, experiments, and systems built on real data.
M.S. Data Science student at NYU. I work at the intersection of statistics, machine learning, and engineering—benchmarks, inference, and full-stack tools.
Graduate
NYU · Data Science
Expected
May 2027
Undergraduate
John Jay · CS & Security
About
Brooklyn-based · curious about small data & big models
I am an M.S. Data Science student at New York University (expected May 2027) with a B.S. in Computer Science & Information Security and a mathematics minor from John Jay College (CUNY), May 2025.
Recent work spans distributed recommender systems on Spark, controlled ML benchmarks for building energy, and a multimodal math benchmark (MathLABS). Outside the terminal: volleyball, pickleball, and trails.
Stack
Tools I reach for
Ordered for how I actually build: from raw tables to deployed interfaces.
Python · pandas · scikit-learn
Spark · MLlib · ALS · MinHash LSH
Big data · HDFS · Parquet · Dataproc
ML · trees · LSTM · ablation
Stats · ranking metrics · evaluation
SQL · MySQL · SQLite
MongoDB · JSON pipelines
APIs · REST · model routers
Streamlit · Django
Git · Linux · VS Code
Projects
Selected work
Aligned with the Fall 2026 résumé: big-data recommender systems, controlled ML experiments, and multimodal evaluation. Older builds live in the archive.
Predicting Building Energy Consumption
10 methods across 4 families on 20.2M rows with fixed features/splits. Feature ablation: lag/rolling drop hurt trees 400–615% vs. 8–12% for linear models. Global LSTM (168h lookback) over 970 meters ranked third overall.
MathLABS
400+ visual discrete-math items, MongoDB + Hugging Face pipeline, hybrid evaluator with Gemini and OpenRouter—nine VLMs on math MCQs.
BMCConnect
Curriculum search, live chat, and analytics for ~1,500 CIS students.
More projects & lab archive → — RateMyProfessor Analytics, PlantID, phishing NLP, Emotionfy, and coursework.
Path
Education & experience
Education
M.S. Data Science · NYU
New York, NY · expected May 2027
Coursework: intro DS, optimization & linear algebra, learning from small data, ML, big data, advanced Python, visualization for ML, NLP.
B.S. CS & Information Security · John Jay (CUNY)
Minor in Mathematics · May 2025
OOP (C++), data structures & algorithms, probability & statistics.
Experience & programs
College Assistant · BMCC
Sep 2023 – Dec 2023
CS 101 lectures and tutoring for 20 students.
Technology Learning Community Intern · BMCC
May 2023 – Jul 2023
Curriculum search, chat, and dashboards for CIS students.
CodePath · Apprenticeship
Jun 2025 – Aug 2025
Web dev & technical interview prep tracks.
CUNY Tech Prep · Data Science Fellow
Jul 2024 – May 2025
EDA through ML with Python, Jupyter, pandas, TensorFlow, scikit-learn.
Contact
Let’s talk data
Email, GitHub, or LinkedIn—best for collaborations, internships, or swapping project ideas.