Fraud Data Science
Summary
Fraud-focused Data Scientist at a fintech lending startup building and validating ML classification models on large transactional and behavioral datasets to detect anomalies and automate fraud decisioning. Core stack: Python, SQL, AWS.
About Us:
Vana is a fintech company founded in 2018, specializing in providing quick and accessible personal loans through a user-friendly Android application. By leveraging non-traditional data sources and machine learning algorithms, Vana offers consumer financing solutions to individuals who may not have access to traditional credit facilities.
Vana is committed to addressing the significant market demand for quick and affordable credit access in Latin America, aiming to empower users by providing convenient loans that can be applied for in minutes through their Android app.
As a Data Scientist focused on fraud prevention, you will play a key role in designing and developing models that help identify anomalous behavior, suspicious transactions, and fraud-related patterns. You’ll work with large datasets and apply supervised and unsupervised machine learning techniques to strengthen our fraud detection systems and automate decision-making.
This role requires a strong understanding of imbalanced classification problems, solid feature engineering skills, and the ability to adapt models to evolving user behavior and new fraud strategies. Experience in fintech, banking, insurance, or other industries with a strong focus on fraud prevention and automated controls is desirable.
Responsibilities
- Data analysis and modeling: Explore large-scale transactional and behavioral datasets to uncover patterns associated with fraud.
- Model development and validation: Build and validate classification models using ML techniques tailored to imbalanced problems (fraud vs. non-fraud).
- Feature engineering: Create derived variables that enhance model performance and generalization while avoiding overfitting.
- Cross-functional collaboration: Work with engineering, product, and operations teams to ensure seamless integration of models into decision flows.
- Monitoring and iteration: Track model performance in production and iterate based on behavioral changes, fraud trends, or strategy shifts.
- Research and innovation: Stay up to date on cutting-edge ML techniques for fraud detection in digital transactional environments.
Requirements:
- Degree in Data Science, Statistics, Mathematics, Computer Science or a related field with strong programming skills [MUST]
- 1-2 years of experience in Data Science, Data/Business Analytics (with ML knowledge) [MUST]
- 2+ years of Python and SQL experience [MUST]
- Knowledge of fraud detection, anomaly detection, or modeling with imbalanced datasets [MUST]
- Industry background in fintech, insurance, or banking is valued but not required [DESIRABLE]
- AWS Services knowledge is a Plus [DESIRABLE]
- Strong analytical skills and a problem-solving mindset, with the ability to extract actionable insights from data [MUST]
- Understanding of consumer behavior, alternative data sources, and digital lending platforms [DESIRABLE]