- Applied statistical techniques on GIS data to evaluate the impact of meteorological variables on crop yield (e.g., explain yield variability using drought index).- Created models using Machine Learning techniques (e.g., logistic regression, support vector machine, random forest, KNN) to locally improve leaf wetness classification (i.e., increased accuracy and true positive rate).

Working with data analysis to bring solutions related to the auto insurance business:- Developed and updated statistical models to estimate auto risk premium (e.g., frequency and severity of car accident) using GLM techniques (Poisson and Gama distribution). - Developed and updated churn models for to estimate the probability of a client renew its insurance policy. Optimize the model to increase profitability (e.g., increase profitability constrained to a maximum of 20% of churn).-… Show more Working with data analysis to bring solutions related to the auto insurance business:- Developed and updated statistical models to estimate auto risk premium (e.g., frequency and severity of car accident) using GLM techniques (Poisson and Gama distribution). - Developed and updated churn models for to estimate the probability of a client renew its insurance policy. Optimize the model to increase profitability (e.g., increase profitability constrained to a maximum of 20% of churn).- Time series analysis to identify trends on auto insurance risk in order to adjust risk premium. - GIS analysis: Improved spatial information for auto insurance business reducing error by region.- Provide assistant to priority projects for the company – which included suggesting any new approach or analysis that could add value to the company. Show less

- Created and updated sales dashboard (KPIs) to manage the performance of sales team with different views. - Developed client segmentation as part of CRM for marketing strategies (Cluster analysis – Unsupervised learning: K-Means and Hierarchical). - Modeled churn using classification techniques (e.g., logistic regression). Implemented strategies to reduce churn via call center (discount decision tree). Responsible for updating and sending daily churn report.

- Manipulated large datasets (millions of credit card transactions) using SQL to manipulate data (e.g., aggregating, merging) in order to create datasets for analysis. - Time series analysis to forecasts clients’ revenue to identify trends for marketing strategies. - Modeling transactions: Part of the team responsible for model transaction fraud risk (logistic regression).