Description:

The PayU Intelligence Team is hiring a Data Scientist to model complex transactional problems, discover hidden insights, and construct predictive systems integrated with live fintech services. This role has a strong focus on advanced data mining, applying graph databases/analytics, and executing multivariate testing. You will design data augmentation pipelines, implement automated real-time anomaly detection frameworks, and architect end-to-end deep learning models to identify patterns over massive credit, fraud, and payment datasets.

Key Technologies:

Python, R, SQL, Graph Databases, Neo4j, TensorFlow, PyTorch, Keras, Scikit-learn, Pandas, Hadoop MapReduce, Hive, Pig, Java, SAS

Requirements:

• Bachelor's degree in Mathematics, Statistics, Computer Science, or a related quantitative field. A Master's or Ph.D. degree is highly preferred.
• 2–6 years of relevant quantitative and qualitative research, data science, and analytics experience.
• Extensive knowledge of statistical techniques, quantitative analytics, forecasting, multivariate testing, and complex optimization algorithms. Ability to quickly grasp mathematical concepts down to the core theorems to solve original abstract issues.
• Advanced proficiency deploying classic ML (SVM, GLM, boosting, random forest), Deep Learning structures (CNN, RNN, LSTM, attention models), Graph Neural Networks/models, and Reinforcement Learning architectures.
• Hands-on experience with open-source engineering packages such as Keras, TensorFlow, PyTorch, scikit-learn, and Pandas.
• Strong programming skills across big data frameworks (Hadoop MapReduce, etc.), statistical suites (R, Python, SAS), and data query languages (SQL, Hive, Pig). Familiarity with basic distributed computing principles.
• Define preprocessing, advanced feature engineering, data augmentation pipelines, hyperparameter tuning, error analysis strategies, and automated anomaly tracking systems.
• Exceptional problem-solving and business modeling skills to extract actionable solutions out of loosely defined, unstructured business problem statements.