David Mantilla Email & Phone Number

New York City Metropolitan Area

São Paulo, Sp, Br

Head of Data Science for Global Lending at NubankLed a 30+ person team comprising data scientists, machine learning engineers, and analytical engineers committed to delivering world-class machine learning solutions that power our growing lending business.* Accountable for all credit risk models and price sensitivity models used by Global Lending, which helped achieve 76% YOY portfolio growth in personal loans in 2023 for Nubank’s 78M active customers* Presented to C-suite management team (e.g., President, CMO) on Lending’s credit infrastructure projects, ensuring buy-in on long-term initiatives, investments, and needed improvements* Recognized for strong people management and talent development skills, receiving consistently high engagement and team satisfaction scores for global team

Mclean, Va, Us

* Led team (as hands-on data science/MLE manager) in building real-time, transaction-level machine learning underwriting capability (with under 150ms latency) powering Capital One's new Cash Plus card, a pay-in-full, no-preset spending limit charge card, launched Aug 2021* Earned Circle of Excellence award, Capital One’s highest internal recognition, for delivering innovative capabilities around real-time credit risk models powering Small Business Card's future growth* Led Small Business Card’s Customer Management Behavioral Modeling Data Science team to build models leveraging bureau, transaction, and internal performance data to boost program eligibility and improve credit line sloping* Designed and implemented modeling infrastructure leveraging pySpark, H2O, Snowflake, Redshift, Teradata, AWS EC2 and EMR, python, and CI/CD tools * Trained and mentored statisticians-turned-data-scientists on software engineering best practices, object oriented design principles, and machine learning methods

Mclean, Va, Us

* Implemented machine learning components of Second Look, a data service that identifies unusual credit card transactions for our customers, using scikit-learn + Hadoop Map Reduce on 35M+ customer base; went live in 2016* Managed an "inner-source” software development project based on Spotify's luigi framework and used by several big data projects at Capital One; increased internal contributor base by 200% in 9 months, and helped Capital One embrace open source software throughout the enterprise (marking a big IT policy change at the time)* Provided operational support to Enhanced Transactions, a Hadoop map reduce-based entity resolution service that matches transaction records to a merchant directory listings, including co-authored model white paper, and performed training data labeling, model performance monitoring

New York, Ny, Us

Led effort to build out robust analytics and business intelligence capabilities at leading international ecommerce company as a technical, hands-on manager, leader, and solution architect for client-facing multi-tenant reporting solution.• Architected the multi-tenant reporting solution, including networking, server hardware, data flow, ETL and data warehouse (star schema) and OLAP cube design, enabling its 200+ clients and internal staff access to self-service reports and dashboards• Evangelized use of Pentaho, Tableau, open source technologies, and Vertica database within the organization, saving the company over $250,000 in licensing costs while ensuring low on-going support costs

Bachelor Of Arts - Ba, Computer Science

Master Of Business Administration - Mba