Data Scientist - Client Classification & Risk Control

About the role

Hytech, a management consulting firm based in Australia and Singapore, is looking for a Data Scientist to drive its machine learning and deep learning research agenda for a client-classification engine and adjacent transaction risk control and anti-fraud scenarios. The Data Scientist owns the end-to-end experiment loop, from hypothesis through backtest, calibration and monitored rollout, so that routing and control decisions improve broker profitability, stability and abuse prevention.

Key responsibilities

• Own modeling research execution: turn business and user questions into measurable data solutions and work with analysts and engineers to build, test and iterate.
• Explore and apply ML/DL models to solve practical integrity and classification problems, such as user behavior modeling, fraud/anomaly detection, reputation/risk scoring or abuse prevention.
• Design fair, time-aware evaluation strategies, run robust backtesting and validation, and report findings and next steps clearly.
• Partner with engineers to ship models safely, monitor their health and set up retraining or recalibration triggers.
• Write concise design documents and experiment summaries understandable by technical and business stakeholders.

Required profile

• 5+ years in Data Science, ML or DL, with a track record of solving practical business or user problems.
• Solid experiment design fundamentals: careful validation, fair metrics and awareness of data pitfalls.
• Effective communicator who can summarise findings for technical and non-technical partners.

Required skills

• Strong Python and SQL, with clean, reproducible analysis code in notebooks or scripts.
• Hands-on with at least one ML or DL library (scikit-learn, PyTorch, TensorFlow).
• Nice to have: deep learning and time-series models (LSTM, GRU, Transformers), representation learning, handling noisy or partially labeled data, large-scale experimentation with Spark/Databricks/MLflow, and experience in fraud, personalization, search, ads or growth analytics.