Problem: Improve the company’s operating result and, at the same time, the performance in credit recovery for large customers.
Data sources: Text files (CSV), spreadsheets (MS Excel), relational databases of CRM systems; database of individuals with more than 50 million lines (dozens of transactions for each line, with variations according to the portfolio).
Solution: Predictive analysis models to define score to payment propensity; cluster definition; visual analysis models to support all levels of operation and clients.
Challenge: Prepare data (data blending, data enrichment) from various sources and structures; insert data from external sources (data suppliers for enrichment and public data); introduce cultural analytics in the area of operations and throughout the organization.
