>_ TIMESTAMP: Oct 2025

Engineered feature vectors (HOG, GLCM) to isolate texture variance, distinguishing signal from background Quantified feature importance via Random Forest ranking, reducing 750+ dimensions to maximize signal fidelity Architected Active Learning loops via Uncertainty Sampling, utilizing SMOTE to resolve class imbalance Optimized XGBoost hyperparameters via GridSearchCV, achieving 0.85 F1 on highly skewed datasets

>_ TIMESTAMP: Jan 2025

Engaged in a rigorous exploration of mathematical finance, focusing on the theoretical and computational frameworks underlying derivative pricing. Covered core topics including interest rates, bond mathematics, portfolio theory, forwards/futures pricing, hedging strategies and options valuation using both discrete (binomial) and continuous-time (Black-Scholes) models. Implemented complete simulation workflows in Python to evaluate strategies such as protective puts, covered calls and delta hedging. Applied optimization techniques (e.g., Sharpe Ratio maximization) for portfolio design.

>_ TIMESTAMP: Jan 2025

Designed and implemented a modular Python-based framework for developing, simulating, and optimizing algorithmic trading strategies in cryptocurrency markets. This end-to-end system comprises three core modules: Historical Data Acquisition: Integrated with Delta Exchange API to fetch granular tick-level and order book data for futures and options. Event-Driven Simulation Engine: Built a high-fidelity backtester that models slippage, transaction costs, and execution mechanics. Supports both directional and complex derivatives strategies. Strategy Optimization: Leveraged Multi-Armed Bandit (MAB) algorithms to efficiently navigate the parameter search space, improving convergence and robustness of strategies.

>_ TIMESTAMP: Jan 2025

AI voice generation using Mamba and State Space Models, emphasizing linear scaling relative to quadratic attention.