Research data scientist / research software engineer
Data scientist and research software engineer working on reliable ML, statistical pipelines, and AI systems that can be evaluated. My path has moved through software, IoT analytics, biomedical research, and finance, and I am now building toward deeper statistical work at the University of Geneva.
About
I came into data science through applied work rather than toy problems. I worked with chemical-data software, IoT sensor networks, neurological research data, and credit-risk reporting. Those environments taught me that useful ML is rarely just the model. It is the data pipeline, the assumptions, the validation checks, the interface, and the honesty about uncertainty. That is the direction I am taking into statistics and research software.
I like work where the implementation and the reasoning have to hold together. A good project should be easy to inspect, possible to reproduce, and honest about what is known.
The projects I want to keep building sit between statistical thinking and practical systems. I use them to become stronger in software development while building a deeper foundation in research and statistics, so I can approach new technologies with enough breadth to understand the whole system and enough depth to apply them carefully.
Project portfolio
A compact set of projects ordered by technical scope, maturity, and relevance to my current research and engineering direction.
Local-first document QA backend with ingestion, hybrid dense/BM25 retrieval, reranking, Ollama generation, citation filtering, workspace isolation, and evaluation.
Reproducible systemic-risk monitoring pipeline for the US, Germany, Spain, and the UK, built from open macro-financial, market, leverage, and uncertainty indicators.
FastAPI chatbot application combining sentence-transformer embeddings, calibrated logistic regression intent classification, templates, Docker, and a retraining endpoint.
Background
Data Science Intern
Reworked monthly credit reporting into a weekly workflow, adapting data sources, validation checks, execution logic, and BI visualisations for risk monitoring.
Data Scientist
Owned IoT data workflows for thousands of sensors and hundreds of hubs. Built ETL, PostgreSQL/Grafana analytics, backend processing, and regression-based anomaly detection.
Data Science Intern
Harmonised neurological research datasets, prepared APOE biomarker data, and supported exploratory and Kaplan-Meier analysis in Python, R, and remote research environments.
Software Developer
Optimised SQL queries and refactored database schemas, improving ETL runtime and maintainability for data-processing workflows.
Education
University of Geneva
Open University of Catalonia
Escuelas Universitarias Gimbernat
Tooling
Python, SQL, R, SAS, Bash, MATLAB, Swift, JavaScript, Go basics.
Regression, anomaly detection, supervised and unsupervised learning, NLP, embeddings, RAG, LLM systems.
EDA, Bayesian inference, PCA, survival analysis, calibration, conformal prediction, credit-risk metrics.
FastAPI, PostgreSQL, Docker, Git, ETL pipelines, validation checks, Grafana, Linux/VNC, AWS.
Contact