SPECIALIZING IN BUSINESS AUTOMATION & AI IMPLEMENTATION
Former private equity research analyst turned AI consultant and data scientist. Helps businesses streamline operations by automating tedious processes and identifying the right AI tools for specific challenges. Focus: making work more efficient, reducing manual tasks, and implementing practical AI solutions that actually work.
Based in Montreal, Canada
Expert in business process automation
Open to collaboration on AI implementation
Passionate about practical AI solutions
Advanced Machine Learning Framework for Computational Fluid Dynamics
Revolutionary approach combining computational fluid dynamics with machine learning, achieving 99.54% accuracy in drag coefficient prediction across multiple flow regimes. Features physics-informed neural network architecture, logarithmic Reynolds number transformation, and domain-specific feature engineering. Demonstrates training efficiency (~10 seconds on CPU) while maintaining physical interpretability - critical for engineering applications requiring model reliability and understanding.
Advanced Reduced-Order Modeling with Non-Intrusive Operator Inference
Comprehensive implementation of the "Lift & Learn" methodology for learning reduced-order models of the FitzHugh-Nagumo system. Features complete 5-phase implementation (Data → Lifting → POD → Operator Inference → Validation), noise robustness analysis, and interactive visualizations. Achieves <1e-3 error with 100x speedup, demonstrating practical ROM deployment for nonlinear PDEs with professional-quality documentation and extensible framework design.
Statistical Physics Investigation of Perceptron Storage Capacity
Comprehensive numerical investigation revealing a mathematical puzzle where a flawed derivation yields correct results. Features first-ever 3D visualization of Gardner phase space, universal formula validation (α_c = 1/(κ²+1)), and demonstrates hidden symmetries in high-dimensional statistical physics. Achieved R² = 0.9799 correlation between theory and simulation, providing insights into generalization bounds and phase transitions in neural networks.
Revolutionary Physics-Informed Multi-Objective Machine Learning Framework
Breakthrough computational photonics framework combining multi-objective NSGA-III optimization with physics-informed machine learning. Features automated discovery capabilities, multi-fidelity intelligence, and 13+ physics features for topological photonic crystal design. First framework to simultaneously optimize Q-factor, robustness, bandgap, and mode volume with automated design rule extraction. Establishes new performance benchmarks in computational photonics research.
Group Robust Policy Optimization for Ethical Medical Resource Allocation
Novel reinforcement learning algorithm extending PPO with explicit fairness constraints. Achieved 25% fairness improvement with statistical significance while maintaining operational efficiency across multiple hospital scenarios.
Advanced Mathematical Toolkit for Nonlinear Dynamical Systems
Comprehensive Python framework for analyzing chaotic dynamical systems using advanced mathematical techniques. Features automated parameter optimization, fractal dimension estimation, and publication-quality scientific visualizations.
Neural Physics Approach to Financial Market State Classification
Novel application of Hopfield networks to financial markets, treating market regimes as stable attractors in the network's energy landscape. Uses content-addressable memory for pattern recognition and regime classification based on technical indicators.
Entropy-Adaptive Cybernetic Threshold Optimizer for Dynamic Financial Risk Management
Advanced financial risk management system that dynamically adapts risk thresholds based on market complexity measures. Leverages information theory, statistical modeling, and cybernetic control systems to outperform static risk models.
Advanced Quantitative Finance Library for Institutional-Grade Analytics
Comprehensive quantitative finance library implementing advanced mathematical models used by hedge funds and investment banks. Features portfolio optimization, VaR/CVaR risk metrics, Black-Scholes options pricing, market regime detection, and real-time data integration.
Optimal Feedback Control for Pathological Biological Oscillations
Advanced control engineering project demonstrating Linear Quadratic Regulator techniques for stabilizing biological oscillators. Features nonlinear system linearization, optimal control design, and therapeutic applications for circadian rhythm disorders and metabolic network stabilization.
Q-Learning Agent for Stochastic Environment Navigation
Reinforcement learning demonstration featuring a Q-learning agent that learns optimal navigation through uncertain environments. Handles movement unpredictability, obstacle avoidance, and policy adaptation under different levels of environmental uncertainty with comprehensive visualization of learning progress.
Interested in collaborating on business automation or AI implementation projects? Let's discuss how we can work together to streamline your operations and build practical AI solutions.