I am a dual-degree student pursuing an MSc in Artificial Intelligence at UQAC (Canada) and a Master of Engineering in Computer Science at Télécom Saint-Étienne (France). Coming from an intensive mathematics and physics background (Classes Préparatoires MPSI/PSI & Lycée Henri-IV), my academic core lies in Machine Learning and High-Performance Software Engineering.
Driven by a strong personal interest in financial markets, I am actively applying my technical background to Quantitative Finance through self-study and practical projects. My current focus is on building low-latency C++ systems and applying Deep Learning / Reinforcement Learning to market data.
- 🎓 Currently learning: Advanced Stochastic Calculus, Options Pricing (Black-Scholes framework), and C++ Memory Optimization.
- 🎯 Looking for: A 6-month Quantitative Research / Machine Learning Engineering Internship starting in September 2026.
- 📫 How to reach me: arnoldb2016@gmail.com
Quantitative & AI Research:
High-Performance Engineering & Backend:
Tools & Infra:
- ⚡ Low Latency Deep Learning Inference Engine Designed and developed a C++ inference engine from scratch to replace classical stochastic methods (Monte-Carlo) with a pre-trained Deep Neural Network for derivatives pricing. Heavy focus on memory optimization and bypassing Python framework overhead.
- 📈 Automated Trading Agent (DQN) Implemented a Deep Q-Network (Reinforcement Learning) environment from scratch to optimize trading strategies. Backtested on AAPL historical time-series data using Yahoo Finance.
- 📰 Multimodal Algorithmic Trading System Developed a Python algorithm correlating technical market indicators with fundamental analysis through NLP (Sentence-BERT sentiment analysis on live news and Twitter streams).
- 🛡️ LLM & GenAI for Cyberattack Detection Applied research project (in partnership with ReachFive) combining LLaMA 3, LogBERT, and Reinforcement Learning for real-time anomaly detection in server logs.
- 👤 Facial Recognition via Eigenfaces (PCA) Algorithmic implementation of Principal Component Analysis and SVD to create an orthonormal basis of eigenvectors for biometric classification.