• Developed OOP risk system with Python and SQL, useful to implement backtests and quant strategies. Focused on the implementation of options, futures, and forwards, with particular attention to FX and interests rates products on CME and BMF. Currently, it saves 15 hours/week of a senior analyst in exposure calculation and adjustment of pension funds positions.

• Created a VaR simulator in Excel and Django app using API in VBA, Python, SQL Server, and Azure functions for the hedge fund portfolio managers. The software enable the calculation of VaR for a portfolio using multiple timeframes, metrics and methods such Skrinkage of the correlations and EWMA for returns.

• Developed a project to define the probability of default of companies in credit cards using XGBoost and SHAP to assure a good relationship between variables.

• Created time-series models to understand and predict how shifts in macroeconomic factors affect the default rate of each homogeneous credit risk group.

• Created a model to cluster new clients into different groups (use of K-Means with T-SNE, PCA, silhouette and elbow). With a priori knowledge, created “personas” for each cluster in order to separate clients with high capacity of payment and income from those without them. Deployed model to give information to (i) the credit politics area to assign different processes for new clients with better credit characteristics, and (ii) the commercial and investment sector.