Matheus

MATHEUS KAO LinkedIn | GitHub PROFESSIONAL SUMMARY Senior Data Scientist with over 5 years of experience in major corporations (Itaú, Mercado Livre, Braskem), specializing in high-complexity risk models. Expert in Predictive Modeling (Credit and Churn) with a solid track record of ensuring technical robustness and governance for AI solutions. Strategic background in challenging technical assumptions and implementing monitoring frameworks to mitigate risk and optimize decision-making under uncertainty. Proven expertise in stakeholder management, translating complex technical risks into strategic recommendations for senior leadership. TECHNICAL SKILLS • Languages & Tools: Python, SQL, Google BigQuery. • Advanced Analytics: Predictive Modeling, Causal Inference, Elasticity Modeling, Segmentation, Revenue Management, Tableau, Looker Studio, Power BI, KPI Frameworks, Incrementality Testing, ROI/ROAS Analysis, Attribution Modeling, Brand Lift Insights • Product Science: Teste A/B Testing, Design of Experiments, Hypothesis Testing. • AI & Machine Learning: GenAI (Multi-agents), Tensorflow, Keras, PyTorch, Reinforcement Learning, XGBoost, CNN, RAG (Retrieval-Augmented Generation), Embeddings, Document Chunking, Semantic Search, Prompt Engineering, MCP. • Data & Infrastructure: PostgreSQL, Elasticsearch, Kafka, Apache Airflow, Spark/ Hadoop. • Cloud & DevOps: AWS Machine Learning Specialty (SageMaker, ECS,S3, Glue, Kubernetes), Docker, CI/CD (Github Actions), Jenkins, MLFlow. • Business: Demand Planning, Integrated Business Planning (IBP), Pricing Strategy, Portfolio Strategy, Financial Planning, Stakeholder Management, ROI Analysis. • Growth & Marketing: Conversion Optimization, Churn Analysis, LTV, Acquisition Analytics, SEO/Growth Insights, Credit Scoring. • Testing & Practices: Agile/Scrum, Technical Leadership, Jira, Miro, Monday. PROFESSIONAL EXPERIENCE Senior Data Scientist Marketplace Mercado Livre 08/2024 Ongoing • Causal Inference & Strategic Decisions: Led impact analyses using Causal Inference to isolate the effects of product changes, providing independent technical challenges to growth models to ensure the reliability of highimpact decisions. • Model as a Product (MaaP): Developed scalable Machine Learning architectures treated as internal data products, focusing on governance, user adoption, and large-scale extensibility. • Strategic Partnership: Partnered with C-level executives to synchronize predictive models with Integrated Business Planning (IBP), ensuring model outputs remained aligned with global risk appetite and business goals. Data Scientist Itaú & Unibanco 01/2024 - 08/2025 • Revenue & Churn Optimization: Implemented XGBoost and Causal Inference models to identify root causes of disengagement, establishing feedback loops that recovered BRL 300M/year by increasing model precision and enhancing decision-making tools. • Product Behavior & Personalization: Architected Generative AI and Reinforcement Learning solutions to personalize user experience, driving financial product conversion and platform engagement. • Financial Risk & Credit: Structured analytical initiatives from business problem framing to the delivery of Generative AI solutions for hyper-personalized offerings. • Stakeholder Influence: Translated complex technical insights into strategic recommendations for directors, influencing the product roadmap and long-term retention strategies. MATHEUS KAO LinkedIn | GitHub • Strategic KPI Tracking: Defined and monitored strategic KPIs for executive boards, translating technical data into actionable narratives to optimize Customer Lifetime Value (LTV). Global Data & Analytics Braskem 01/2023 - 02/2024 • Executive Strategy Support: Developed KPI dashboards and expenditure forecasting models to support executive decision-making across complex global portfolios. • Automation & Scalability: Led automation and modeling projects in the US, Mexico, and Europe, ensuring data pipeline reliability and the integrity of strategic reports for executive leadership. • Advanced Modeling: Developed classification and regression models for portfolio optimization, focusing on model stability and consistent performance across complex global datasets. • Data Pipeline Collaboration: Utilized Python to optimize workflows, establishing technical playbooks for automated reporting and continuous model monitoring. EDUCATION MBA Data Science Bachelors Chemistry FIAP University of São Paulo 02/2024 - 2026 02/2019 - 12/2023 PROJECTS • Specialization in Machine Learning & Data Mining (2024): SHIFT technology & FIAP. • EY (Ernst & Young) Data Challenge 2024: Developed a project utilizing Convolutional Neural Networks (CNNs) to predict and measure the impact of climate change.