• Built a futures volume forecasting engine in Python/kdb+ predicting time-of-day liquidity profiles, using SMA/EWMA calculations on historical profiles to identify seasonal and regime-driven shifts across contracts.
• Constructed a shape-based FX volume regression model learning hourly liquidity patterns from CLS data labels, using 1-day lagged spread/quote/RV EWMA anchor features across multiple spans to generate NDF volume profiles.
• Achieved 0.60 pooled R² (0.56 mean per-pair) in volume share-space — 47% above a per-pair quote-EWMA benchmark (0.38 R²) and within 15% of a per-pair volume-EWMA benchmark (0.66 R²).
• Backtested scheduling in a tick-level simulator, reducing realized slippage by ~13% and holding participation under 2.5% with volume-profile VWAP execution versus benchmark TWAP scheduling.

Jun — Sep 2026

• Engineered a Black-Scholes options pricing model and an L3 depth order book with NumPy and Pandas, processing raw ITCH protocol intraday market data with 50k+ trades and 500k+ orders.
• Developed a modular backtesting framework integrated with a ClickHouse DB to evaluate passive liquidity (buy-wall) and active momentum strategies with low-latency data retrieval against the ITCH order book.

Aug — Sep 2025

• Built an Oracle DB and MySQL pipeline to preprocess 1M+ financial transactions across 1,000+ accounts.
• Developed a Logistic Regression fraud detection model via scikit-learn, achieving 81% accuracy.

Aug — Sep 2023

• Awarded the internship after winning Robert College's annual elevator pitch competition.
• Scouted 100+ startups via startups.watch and delivered a fintech growth projection report using Python data analysis.
• Developed understanding of SaaS products by attending the SaaStanbul conference.

Jul — Aug 2023