I'm a Mathematics & Computer Science student at NYU Abu Dhabi building projects at the intersection of machine learning, quantitative finance, and software engineering.

My background in competitive mathematics and programming shapes how I work: I like turning mathematical ideas into clean, reproducible systems — from financial modeling and credit-risk audits to deep learning pipelines, LLM fine-tuning, and production web applications.

• Machine Learning Systems — deep learning pipelines, model evaluation, fine-tuning, ablation studies, and reproducible experiments.
• Quantitative Finance & Statistical Modeling — factor models, regression, backtesting infrastructure, credit-risk analysis, and trading strategy research.
• Software Engineering Projects — full-stack applications, production websites, clean APIs, testing, CI, and deployment-ready project structure.
• AI Research Prototypes — LLM fine-tuning, optimization methods, model diagnostics, and applied research notebooks.

SmolLM LoRA Fine-Tuning
Parameter-efficient fine-tuning of SmolLM-135M using LoRA, with rank ablation, W&B experiment tracking, perplexity evaluation, and qualitative failure-mode analysis.

Vehicle Trajectory Behavior Classifier
CNN-LSTM pipeline for classifying driving behavior from speed and position time-series data, focused on sequence modeling and mobility analytics.

Reactiva Perú Credit-Risk Audit
Replication and methodological audit of credit-risk ML models, identifying target leakage through statistical testing and Random Forest validation.

Trading Strategy ML
Machine learning research pipeline for financial time-series modeling, strategy training, evaluation, and reproducible experimentation.

Hedge Fund Returns Analysis
Ridge regression and Fama-French factor analysis for hedge fund return modeling, including bias-variance analysis and feature engineering.

AcademicHub
Production Next.js website for a Moldovan mathematics school preparing 100+ active students for national exams, with enrollment flows, SEO, and lead-management features.

Halma Minimax AI
Python game AI implementing Minimax and Alpha-Beta pruning for strategic decision-making in Halma.

Languages: Python, TypeScript, C++, SQL
ML/Data: PyTorch, scikit-learn, pandas, NumPy, Jupyter, W&B
Software: Next.js, React, Tailwind CSS, FastAPI, Docker, GitHub Actions
Finance/Stats: regression, factor models, time-series analysis, backtesting, risk modeling
Interests: ML engineering, AI agents, quantitative trading, financial modeling, optimization, applied AI research