Fraud Data Scientist

Job Title: Fraud Data Scientist

Experience Level: 2–5 Years

Location: Bangalore, Gurugram, Hyderabad

Employment Type: Full-Time

About Straive:

Straive is a market leading Content and Data Technology company providing data services, subject matter expertise, & technology solutions to multiple domains. Data Analytics & Al Solutions, Data Al Powered Operations and Education & Learning form the core pillars of the company’s long-term vision. The company is a specialized solutions provider to business information providers in finance, insurance, legal, real estate, life sciences and logistics. Straive continues to be the leading content services provider to research and education publishers. Data Analytics & Al Services: Our Data Solutions business has become critical to our client's success. We use technology and Al with human experts-in loop to create data assets that our clients use to power their data products and their end customers' workflows. As our clients expect us to become their future fit Analytics and Al partner, they look to us for help in building data analytics and Al enterprise capabilities for them. With a client-base scoping 30 countries worldwide, Straive’s multi-geographical resource pool is strategically located in eight countries - India, Philippines, USA, Nicaragua, Vietnam, United Kingdom, and the company headquarters in Singapore.

Website: https://www.straive.com/

Role Overview

We are looking for a Data Scientist with 2–4 years of experience to own end-to-end Machine Learning model development for payment fraud prevention. You will design, build, and deploy high-throughput fraud decisioning models and feature pipelines across card transactions, merchant activity, and account takeover scenarios.

Key Responsibilities

• ML Development: Build payment fraud models using XGBoost, LightGBM, and Isolation Forests.
• Feature Pipelines: Build velocity, temporal, and device features using Python and SQL.
• Decision Strategy: Optimize precision-recall trade-offs to balance fraud loss and friction.
• Model Lifecycle: Track production performance (PSI, drift) and manage retraining pipelines.
• Cross-Functional: Translate emerging threat vectors into actionable rules and model signals.

Technical Requirements

• Experience: 2–4 years of hands-on data science experience.
• Core Tech Stack:
• Python: Strong proficiency with pandas, numpy, scikit-learn.
• SQL: Good Hands on capabilities (complex joins, window functions, query optimization on large datasets).
• Machine Learning: Hands-on experience in building ML models.
• Domain Knowledge: Understanding of banking/payments, credit/debit card fraud patterns, transaction monitoring, account takeover (ATO), and velocity-based risk scoring etc.Preferred / Nice-to-Have Skills
• Exposure to cloud platforms (AWS / GCP / Azure) and production deployment frameworks.
• Basic familiarity with MLOps practices, model versioning or real-time scoring engines.