Rafael Barbosa

• Built, deployed and evolved a customer value tiering model across multiple releases, an end-to-end ML solution for predicting future customer value.Improved RMSLE from 2.53 to 1.77 versus the previous baseline, enabling daily scoring at million-customer scale, earlier identification of value changes and material financial impact. Automated the Databricks production workflow with performance moni

• Contributed to building the company's Analytics and Data Engineering foundations, supporting the migration of the data platform from MySQL to Amazon Redshift and developing dimensional models with fact, dimension and aggregate tables within a layered data architecture
• Developed a business-critical aggregation layer in dbt, feeding one of the company's main management dashboards and supporting

Mid-level Data Scientist - UOL CS Project (Universo Online – Content & Services)

• Delivered end-to-end Data Science solutions for advertising, digital content and online products, translating business problems into analytical and modelling approaches
• Developed statistical, Machine Learning and recommendation models to support performance analysis, content relevance and product decision-making

Junior and Mid-Level Data Analyst

• Delivered analytics and statistical modelling initiatives to support marketing performance and strategic business decisions
• Analysed campaign and performance data to identify trends, evaluate results and uncover opportunities for optimisation
• Developed classical statistical, supervised and unsupervised Machine Learning models using R and Python for business

Applied R, Python and data visualisation tools to analyse datasets across academic research projects in different fields.

Economic and Fiscal Information Unit (CIEF) – Revenue Directorate

• Performed data analysis for economic and fiscal datasets, including data cleaning and exploratory analysis
• Developed classical time-series forecasting models in R to support tax revenue forecasts across 13 regional offices
• Produced technical reports in R Markdown and contributed to Power BI dashboards for stakeholder reportin

Ongoing learning in data analysis, descriptive statistics, probability, inference, hypothesis testing, and univariate and multivariate statistical techniques. Use of statistical models, such as linear and logistic regression, and classic time-series models, such as ARIMA and Holt-Winters.

Thesis: A Statistical Approach with Machine Learning for Zone Classification in Water Reservoirs

• Explorato