Rafael A.

Machine Learning engineer

Redsys España • Fulltime

Oct 2025 - Present • 11 mos

Machine learning and generative AI expert working on internal solutions for a corporate group, with a focus on the deployment and maintenance of models in production rather than experimentation alone. I built and operated machine learning and generative AI systems addressing internal business needs, taking models from prototype to reliable, monitored services on Amazon Web Services. My work covered the full deployment lifecycle: containerization and packaging, cloud infrastructure and orchestration, integration with existing internal systems, and ongoing monitoring, retraining and performance maintenance once in production. I also worked with large language models and generative AI to automate internal processes and support corporate workflows. c lProjects were internal and confidential, so specific details cannot be disclosed.

Data Scientist & Machine Learning expert

IMIBIC • Fulltime

Oct 2023 - Oct 2025 • 2 yrs

Machine learning and data expert within a biomedical research team, working with clinical cohorts of over 100,000 patients in hepatitis C and diabetic nephropathy. I developed and validated predictive models for clinical screening and disease staging, including a screening model that reduced the required testing workload by approximately 70% while maintaining case detection. My work covered the full pipeline: cleaning and integration of large clinical datasets, exploratory and statistical analysis, model development and validation, and the reporting of results to clinicians and researchers without a computational background. Primary stack: Python, with scikit-learn and PyTorch for modelling, pandas and NumPy for data processing, and matplotlib for visualization. All code and pipelines were documented to ensure reproducibility.

Data Scientist & Machine Learning expert

Universidad de Córdoba • Fulltime

Dec 2021 - Oct 2023 • 1 yr 10 mos

Data scientist and machine learning expert at the University of Córdoba, working on applied and methodological projects across energy, education and climate, all of which resulted in peer-reviewed scientific publications. For an energy company, I developed wind speed forecasting models for wind farm operation, reducing model complexity to one third of their previous solution while maintaining the same predictive accuracy. I also built ordinal models to predict student performance on a graded scale, reaching 92% AUC by exploiting the natural ordering of the target classes. Together with my team, I co-proposed a new methodology for heat wave reconstruction that improved accuracy by 13% over previous approaches. Alongside these applied projects, I worked on methodological developments in ordinal classification and time series analysis, validating new approaches across a wide range of real-world problems.