Principal Data Scientist (Fraud Modelling)

Location: Cape Town, South Africa

Company

hirify.global operates a commerce partnership marketing platform covering affiliate, influencer, and customer referral partnerships.

What you will do

• Research fraud detection and risk monitoring across attribution, lead quality, click injection, browser extension abuse, brand safety, and creator authenticity.
• Design, prototype, validate, and deploy machine learning models and rule-based systems for fraud detection, partner risk scoring, and compliance workflows.
• Apply graph-based techniques, behavioral clustering, and network analysis to identify coordinated fraud rings and suspicious relationships.
• Own production delivery across ETL, feature engineering, model training, deployment, monitoring, retraining, and drift detection.
• Analyze fraud trends and partner behavior, then translate findings into recommendations for Product, Marketing, Finance, Compliance, and Trust & Safety.
• Build dashboards, communicate model performance and risk metrics, and present technical findings to cross-functional stakeholders.

Requirements

• 5+ years of experience in data science, machine learning, or advanced analytics, including 2+ years focused on fraud detection, risk modeling, or anomaly detection in production.
• Experience building and deploying fraud or risk models using classification, anomaly detection, time-series analysis, or graph-based methods.
• Strong Python and SQL skills, with experience using machine learning libraries such as scikit-learn, XGBoost, or LightGBM.
• Experience with feature engineering, model evaluation, imbalanced datasets, experiment design, and statistical analysis.
• Familiarity with production ML workflows, including versioning, monitoring, A/B testing, and model retraining.
• Bachelor’s degree in a quantitative field such as computer science, statistics, mathematics, or engineering.

Nice to have

• Experience with affiliate marketing, ad tech, e-commerce fraud, browser extension detection, fingerprinting, or identity resolution.
• Experience with graph analytics, network-based fraud detection, privacy-preserving ML, or hybrid rule-ML systems.
• Experience with real-time scoring, REST APIs, streaming pipelines, GCP tools, Databricks, or Spark.

Culture & Benefits

• Flexible working environment with a responsible paid time off policy.
• Up to 12 fully covered therapy or coaching sessions per year, dependent coverage, and monthly gym reimbursement.
• Restricted Stock Units with a three-year vesting schedule, pending Board approval.
• Free Coursera subscription and access to PXA courses.
• Paid parental leave, including 26 weeks for the primary caregiver and 13 weeks for the secondary caregiver.
• Technology stipend for home office setup and a monthly internet allowance.