The experience behind the work.

I help financial institutions and fintechs translate AI requirements into solution designs, working alongside customer teams on technology evaluation and architecture.

These figures describe the scope of customer work, rather than booked revenue or completed deployments.

I led approximately eight project contributors across two global programs, taking responsibility for infrastructure, architectural choices and cloud implementation.

I helped build Cora’s credit-card portfolio from the ground up, engineering the end-to-end underwriting pipeline and the analytical foundations for credit decisions and portfolio monitoring.

At Grupo SBF, I combined recommendation and search modeling, online experimentation and large-scale data processing to improve the digital retail experience.

Scope documented in my OfferFit CV. The record count describes data processed, not unique customers.

Grupo SBF brings together Centauro and Fisia, Nike’s official distributor in Brazil. These brands describe the group context; my employment was with Grupo SBF.

I designed and developed Xper, MLflow-based experimentation software that brought dataset discovery, preparation, feature engineering, tuning and validation into one workflow across credit products.

Built Credit Suite services for eligibility, features, labeling and monitoring, alongside ETL workflows supporting a R$2.5M monthly lending portfolio.

My role was at Noverde in 2019–2021. Dotz acquired Noverde in 2022.

At Banco Votorantim, I combined predictive modeling with practical data engineering for customer retention and write-off collections. My work covered the full path from target definition and feature design to reusable pipelines and model deployment.

I developed Favorites, a graph-based churn model investigating how an influential customer's departure changes risk signals in their close network.

The work drew on message-passing intuition, contagion mathematics and probabilistic graphical models. I used NetworkX and experimented with graph databases and visualization.

My recalled historical ROC-AUC is approximately 0.80–0.85. The original evaluation artifacts are unavailable. The public demonstration uses synthetic data and does not reproduce that result.

Product Management Intern · 2014–2015
Customer Insights Analyst · 2015–2016

While still a Product Management Intern at Monsanto, I initiated a simple NLP approach to producers’ and customers’ NPS responses. I used word counts and text-similarity analysis to surface relevant relationships in open-ended feedback.

My reports helped marketing understand what customers were reporting, including tone, slang and regional language across Monsanto’s corn, soybean and cotton businesses and their brands.

Developed statistical forecasts and customer-insight tools using R, SQL and VBA; analyzed Portuguese NPS feedback and improved reporting and data collection for agricultural decisions.

My role was at Monsanto in 2014–2016. Bayer acquired Monsanto in 2018.