Rayen Gallas rayen003

Rayen Gallas

Business Administration & AI Student | Finance & ML Enthusiast

About Me

Fourth-year double-degree student in Business Administration and AI for Business at Esade Business & Law school.
Passionate about applying AI and ML to solve real-world business problems.
Currently exploring Deep Learning for Financial Analytics and Reinforcement Learning for quantitative trading.

Featured Projects

Here are some of the projects I've worked on:

Project Name	Technologies	Description
Options_RL		Advanced reinforcement learning environment for options trading. Implements differentiable Black-Scholes-Merton pricing with automatic Greek calculation, stochastic volatility, market regimes, and realistic implied volatility extraction. Agent learns dynamic hedging, volatility arbitrage, and synthetic strategies.
LSTM Portfolio Optimizer		Advanced portfolio optimization system using deep learning. Leverages LSTM and Transformer architectures to predict market trends and optimize portfolio weights for better returns.
ML Model Visualization		Streamlit app for interactive ML model training, evaluation, EDA, and experiment tracking.
A/B Testing Visualization		A/B testing app comparing chart effectiveness for data visualization. Features Google Sheets integration, response time measurement, and interactive UI.
AI Study Assistant		A RAG-based AI assistant with math and web browsing capabilities. Built with LangChain, Chroma, and OpenAI, it integrates document retrieval, math problem-solving, and real-time web data fetching.
Agent Factory		Multi-agent AI system framework built with CrewAI. Enables collaborative problem-solving through orchestrated AI agents, featuring automated task distribution and agent communication.
MarketMind		Collaborative AI analysis system using multi-agent architecture. Features specialized agents for market research, trend analysis, and strategic insights generation.
TransformerBard		Advanced NLP model for Shakespeare-style text generation. Implements custom transformer architecture with sophisticated preprocessing and training pipelines.
Real-Time Digit Recognizer		Real-time computer vision system for handwritten digit recognition. Features CNN architecture, webcam integration, and live confidence scoring.
VisionNarrator		Intelligent image description system combining BERT and VGG16 models. Features advanced encoder-decoder architecture for generating natural language descriptions from visual data.
Breast Cancer ML Classifier		Medical diagnosis tool using machine learning. Features PCA-based dimensionality reduction, data preprocessing pipeline, and interactive visualizations for cancer detection.
Docker Web Application		Modern containerized web application showcasing DevOps practices. Implements microservices architecture with Flask backend and PostgreSQL database integration.