BBVA México
Data Scientist – Expert
Leading a team to ensure compliance with regulatory requirements and meet operational goals. Implemented business and expert-defined rules using a blend of supervised and unsupervised models, anomaly detection, and descriptive statistics. Techniques included Isolation Forest, goodness-of-fit tests based on Benford’s Law, and specialized interpretability algorithms.
Designed an algorithm to predict client claim probabilities—reducing the evaluation pool from 1 in 450 to 1 in 34—and identified key factors driving claims to effectively detect fraud.
Deployed large language models to translate internal bank regulations into user-friendly responses, which improved fraud detection accuracy from 19% to 37% while reducing false positives.
BBVA México
Data Scientist – Associate
Classified customers into profiles to prioritize outreach by using tailored classification tools and leveraging the OpenAI API for detailed customer insights. Developed a supervised model (“Best Time to Call”) to identify optimal contact times and integrated this strategy with existing recommendation systems.
This approach identified high-value customer segments (valued at MXN $80,000 per client) while filtering out low-value groups (MXN $102 per client).
BBVA México
Data Scientist – Analyst
Evaluated and adapted machine learning models and ETL processes for forecasting the demand for banknotes and coins across BBVA’s financial subsidiaries. Analyzed market potential, competitive landscape, and key performance indicators to support decisions on deploying or decommissioning ATMs and branches using clustering techniques and data visualization tools.
Developed a decision-support tool that pinpointed optimal growth opportunities, leading to an ATM distribution model that saves MXN 1 million per redistributed ATM annually.
Management Solutions
Business Consultant – Senior Data Scientist
Led a multidisciplinary team to forecast monthly mainframe operations costs using Bayesian linear regression. Developed a machine learning model and interactive dashboard with PySpark and MicroStrategy to predict false clarifications.
Collaborated with the fraud prevention team to prioritize customer clarifications by integrating expert insights, resulting in quarterly savings of up to MXN 1.8 million and increasing daily processing capacity from 25 to 120 cases per analyst.
Management Solutions
Business Consultant – Data Specialist (Assistant II)
Assisted in the migration of corporate tools to Salesforce by translating business rules into front-end development and performing A/B testing on the deployed solutions. Supported the data migration from a data lake to a data warehouse by defining data layouts, formats, and relationships, ensuring information consistency.
This effort culminated in the successful migration of 20,000 million rows of data.
Management Solutions
Business Consultant – Data Specialist (Assistant)
Designed key performance indicators for fraud detection related to excessive sales and prescriptions using principal component analysis and outlier detection techniques. Developed a custom R package to support these efforts and defined KPIs to assess the insured’s willingness and ability to pay using unsupervised learning.
These initiatives led to identifying eight fraud cases per month and the creation of a user-friendly monitoring application, ensuring the health of the insurance portfolio.
