About Me
I am a results-oriented and passionate college junior with a fervent interest in the intersection of data science and finance. My journey has been defined by a commitment to mastering quantitative analysis, machine learning, and financial modeling. I am driven by the challenge of transforming complex datasets into predictive insights that inform investment strategies and drive financial performance.
My experience includes a data analyst internship where I developed my skills in statistical analysis and data visualization. I am currently leading a project focused on algorithmic trading strategies, leveraging data-driven techniques to optimize investment decisions. As a dedicated advocate for STEM education, I volunteer at CoderDojo, where I mentor aspiring programmers and share my passion for technology and quantitative finance.
My Skills
Quantitative Analysis
- Financial Modeling
- Algorithmic Trading
- Risk Management
- Econometrics
- Time Series Analysis
Programming & Databases
- Python (Pandas, NumPy, SciPy)
- R
- SQL
- MATLAB
Technologies
- Machine Learning (Scikit-learn)
- Deep Learning (TensorFlow, PyTorch)
- Big Data (Spark, Hadoop)
- Data Visualization (Tableau, Matplotlib)
Featured Projects
Quantitative Finance Research Project
June 2025
- Built a news sentiment-based trading strategy using VADER and FinBERT to generate market mood signals from financial headlines (NewsAPI, Finviz, Reddit), and getting finance data from yfinance, converting them into actionable buy/sell indicators. Backtested the strategy using a custom engine that accounts for execution costs, and evaluated it with metrics like Sharpe ratio and drawdown.
- Separately implemented foundational quant models such as Black-Scholes, CAPM, Monte Carlo simulations, and Fama-French multi-factor models, and developed portfolio optimization tools using Markowitz and Black-Litterman frameworks.
Adversarial Attacks on Neural Networks
May 2024
- Developed a comprehensive repository showcasing adversarial attacks on neural networks, including Fast Gradient Sign Method (FGSM), Projected Gradient Descent (PGD), and Carlini & Wagner (C&W) Attack, across datasets like MNIST and Fashion-MNIST.
- Explored and documented various attack implementations and their implications for model robustness in critical applications suchs as self-driving and healthcare, emphasizing the importance of resilient AI systems.
Neurosymbolic AI for efficient disease classification
June 2023
- Created a novel approach that combines neural networks and symbolic reasoning techniques to enhance disease prediction and triage management in healthcare settings. By integrating the computational capabilities of neural networks with the logic-based reasoning of symbolic AI, the system aims to improve disease classification accuracy and optimize patient triage processes.
- The neurosymbolic AI system utilizes neural networks to analyze symptom data and identify patterns, while leveraging symbolic reasoning to establish relationships between symptoms and diseases based on predefined rules. The results suggest that this hybrid approach has the potential to significantly enhance healthcare outcomes by providing accurate disease predictions and efficient triage management.