Hi there — I’m Francisco Luís de Avelar Cardoso I'm a Lisbon-based data scientist and machine learning practitioner specializing in time-series forecasting, quantitative modeling, and end-to-end ML systems. I have over two years of hands-on experience building and deploying predictive models across cryptocurrency markets and quantitative trading, with a strong focus on turning theoretically sound models into practical, production-ready solutions. My work spans classical econometrics and modern deep learning, including ARIMA-family models, LSTMs, quantile regression, and hybrid forecasting frameworks implemented in PyTorch, TensorFlow, and Statsmodels. In industry, I’ve designed production LSTM models for multi-asset price forecasting, automated large-scale financial reporting pipelines that reduced manual reconciliation by over 90%, and built ML systems that integrate model training, evaluation, explainability, and deployment. I also developed novel forecasting evaluation metrics (FIS/CER) aimed at measuring financial usefulness rather than pure statistical accuracy, achieving significant improvements in risk-adjusted returns compared to traditional error metrics. My academic background includes an MSc in Mathematics-Economics (ML track) and a Post-Graduate Program in AI & Machine Learning, with deep exposure to numerical methods, stochastic processes, optimization, and statistical learning. I’m particularly interested in roles that sit at the intersection of forecasting, decision-making, and real-world impact, where modeling rigor and engineering discipline matter equally.…Hi there — I’m Francisco Luís de Avelar Cardoso I'm a Lisbon-based data scientist and machine learning practitioner specializing in time-series forecasting, quantitative modeling, and end-to-end ML systems. I have over two years of hands-on experience building and deploying predictive models across cryptocurrency markets and quantitative trading, with a strong focus on turning theoretically sound models into practical, production-ready solutions. My work spans classical econometrics and modern deep learning, including ARIMA-family models, LSTMs, quantile regression, and hybrid forecasting frameworks implemented in PyTorch, TensorFlow, and Statsmodels. In industry, I’ve designed production LSTM models for multi-asset price forecasting, automated large-scale financial reporting pipelines that reduced manual reconciliation by over 90%, and built ML systems that integrate model training, evaluation, explainability, and deployment. I also developed novel forecasting evaluation metrics (FIS/CER) aimed at measuring financial usefulness rather than pure statistical accuracy, achieving significant improvements in risk-adjusted returns compared to traditional error metrics. My academic background includes an MSc in Mathematics-Economics (ML track) and a Post-Graduate Program in AI & Machine Learning, with deep exposure to numerical methods, stochastic processes, optimization, and statistical learning. I’m particularly interested in roles that sit at the intersection of forecasting, decision-making, and real-world impact, where modeling rigor and engineering discipline matter equally.WWWW…