Machine Learning Scientist I - Protect

Requirements:

• Background: You hold a Bachelor’s or Master’s in Computer Science, Data Science, Mathematics, Statistics, or a related quantitative field (or equivalent hands-on experience).
• Core ML Knowledge: Solid grounding in statistical inference, machine learning fundamentals, and predictive modeling (e.g., supervised learning, deep learning, tree-based models like XGBoost/LightGBM).
• Toolkit & Languages: Proficiency in Python and SQL. Familiarity with standard data science libraries (e.g., Pandas, PyTorch, TensorFlow, Scikit-Learn) and distributed computing frameworks (e.g., PySpark) is highly desirable.
• Engineering & MLOps Foundations: Strong interest in clean code, software engineering best practices, and standard MLOps tooling (e.g., MLFlow, Airflow, Git).
• Mindset: You possess a strong curiosity, an experimental mindset, and a passion for learning. You thrive in a collaborative environment where you can launch fast and iterate based on real feedback.

What You'll Be Doing:

• Develop & Train ML Models: Assist in researching, training, and evaluating machine learning models used for real-time risk assessment, anomaly detection, behavioural prediction, and fraud prevention.
• Pipeline Development: Help build and maintain batch and streaming data pipelines to feed online feature stores and analytical workflows.
• Model Lifecycle & Monitoring: Support the continuous improvement of our models by contributing to ongoing automated experiments, validation runs, and real-time performance monitoring.
• Cross-Functional Collaboration: Partner closely with peers in engineering, product, and analytics to translate business requirements into actionable machine learning tasks and merchant-facing risk tools.
• Data Exploration: Analyze large, complex payment datasets to uncover fraud trends, identify patterns, and spot opportunities for ML-driven improvements.
• Experimentation: Participate in model evaluation and experimentation (e.g., A/B testing) to measure model accuracy, efficiency, and real-world business impact.

Nice to Haves:

• No specific nice-to-haves listed.

Perks and Benefits:

• No specific perks and benefits listed.