Jorge Espinoza Abud

• Led a team of data scientists, providing technical guidance and driving risk and fraud detection initiatives.

• Designed, trained, and deployed machine learning models for risk and fraud across all company products, improving bad rates while maintaining approval rates.

• Increased fraud detection related to unwillingness to pay across the portfolio, enhancing risk assessment and mitigation strategies.

• Developed an internal Python library to streamline the data science workflow, reducing development time and ensuring standardization.

• Architected the feature store and built scalable pipelines for training data generation and real-time feature serving in production.

• Established standardized processes and templates for model training, evaluation, monitoring, and feature engineering, significantly reducing time-to-first baseline model and time-to-deployment.

• Defined cross-functional collaboration strategies with business teams, implemented best practices for data science operations, and maintained comprehensive documentation.

• Assessed data providers, conducted proof-of-concept experiments, and recommended high-value data integrations to enhance model performance and business impact.

• Designed and evaluated A/B tests, performing power analysis and statistical tests to determine sample sizes, experiment duration, and data-driven decisions for credit offers and product optimizations.

• Mentored data scientists, providing technical guidance to refine project details while fostering independence and skill development.

• Led data science projects for collections, collaborating with stakeholders to prioritize initiatives, design machine learning solutions, and communicate updates and results effectively.

• Developed early warning detection systems to help the collections team take proactive action at both industry and client levels.

• Built an income estimation model to assign credit lines to clients with incomplete financial data, improving credit decisioning.

• Designed a customer lifetime value (CLV) model to forecast revenue from card clients, supporting long-term profitability strategies.

• Developed a customer churn prediction model, enabling targeted retention strategies for high-risk clients.

• Created a propensity model to estimate customer likelihood of accepting credit line offers, optimizing sales team prioritization.

• Enhanced internal data science tools to improve efficiency in model monitoring and debugging, reducing troubleshooting time.

• Collaborated across three squads—Growth, Risk, and Collections—delivering data-driven solutions tailored to each domain.

• Developed a time series model to analyze credit search behavior at the industry level, identifying trends and market shifts.

• Built a risk estimation model for recurring working capital clients, improving credit decisioning and risk management.

• Designed a collection optimization model to predict the best date to collect payments from clients, enhancing recovery rates and operational efficiency.

• @San pancho (now jellyfish): Conducted descriptive data analysis to identify customer segments and behavioral patterns. Developed a forecasting model to predict trends in customer behavior, optimizing targeting strategies.

• @Teranalytics: Designed a business model for a new product using Monte Carlo simulations. Built a web-based platform to visualize results, enabling clients to make data-driven business decisions based on future projections.

• @BlackInnTech: Developed an ETL pipeline and leveraged NLP models to process unstructured text data, extracting key insights for background checks and driver hiring for a major transportation services app.

• @Sanabria: Assisted with web development and IT support tasks.

• Developed software solutions for the Product Life-Cycle Management (PLM) consulting team, enhancing process efficiency and data management.

• Tech stack: Java, Oracle platform.

• Led the development and maintenance of an ETL system for real estate data, ensuring a reliable data pipeline for company services and analytics.

• Developed NLP models to extract key insights from real estate data, enhancing new features for predictive models in the company.

Data Engineer at Intelimetrica (Mar 2016 - Jul 2016)

• Developed and maintained web scrapers, ensuring efficient data extraction for real estate analysis.

• Contributed to the design phase of the marine propeller production process, optimizing efficiency and performance.

• Worked on the mechanical design and performed stress-strain analysis for a new product, ensuring structural integrity and reliability.