At BRBCARD, I built the Analytics & Business Intelligence function for the acquiring business from scratch, setting its processes, KPIs and roadmap with executive leadership. I also present weekly reporting to the Finance and Operations directorates.
I built a Python churn prediction model that prevented approximately BRL 3 million in TPV loss, and automated validation and reconciliation routines that cut about 15 hours of manual work each week. I applied explainable AI with SHAP to an XGBoost model in a regulated payments environment.
At BB Tecnologia e Serviços, I designed Azure Databricks pipelines on a medallion lakehouse and consolidated 15 heterogeneous sources into analytical datasets. I also built a RAG solution connected to internal HR manuals and an LLM API; earlier, at Neoenergia, I standardized datasets and supported audits and regulatory requests.
