Eloy Celaya LópezData science consultant at Management Solutions, pursuing parallel MSc degrees in AI / Data Science (UCM) and Business Consulting (ICADE). Based in Madrid.
Statistical analysis and ML modelling of residential sale prices across Madrid municipalities. XGBoost outperformed 4 competing models (Linear Regression, SVR, Decision Tree, Random Forest) achieving R²=0.95 — deployed as an interactive Streamlit web app.
Time series analysis of 168 monthly US retail sales observations (FRED/U.S. Census Bureau, 2012–2025). Compared Holt-Winters multiplicative and SARIMA(0,1,2)(2,1,1)[12] models on a 2025 holdout set — Holt-Winters achieved MAPE 1.50% vs SARIMA’s 2.27%, with a trade-off discussion between predictive accuracy and statistical rigour (confidence intervals, residual validation).
Predictive modelling of electoral abstention across 8,117 Spanish municipalities using linear and logistic regression with stepwise variable selection (AIC/BIC). Full preprocessing pipeline: outlier detection, missing value imputation, categorical encoding. Logistic model achieves AUC=0.835 with optimal cut-off at 0.30 (Youden index).
Financial & Credit Risk Consulting practice — risk analytics, data migration, and regulatory transformation for Tier-1 European banks.
Building RDA (Risk Data Aggregation) artifacts and AI agents to automate regulatory data validation and reporting for a major European bank.
Owned end-to-end design of a Power BI risk analytics solution for a leading European banking group — partnering with business analysts to define requirements, sourcing data from Databricks, and building ETL processes (SQL, PySpark) to integrate multiple financial sources into a unified, governed data model with automated KPI calculations.
Supported PMO for a large-scale data migration programme in the financial services sector — tracking milestones, dependencies, and risk across multi-team workstreams, and acting as functional liaison between business stakeholders and IT development teams.
Led and managed a 10-person sales team in Madrid, driving customer engagement and retention over four years while completing undergraduate studies.
Pursuing both in parallel to cover the full spectrum — from technical AI / ML foundations to the business and consulting side of data science.