Description

You will design and implement deterministic detection and transformer-based NLP models for eComms surveillance, including market abuse, insider trading, and conduct risk analysis.

This role is hybrid.

Responsibilities

• Fine-tune transformer models (BERT, RoBERTa, FinBERT) for context-aware detection and develop tokenization, sequence formatting, and label engineering pipelines.
• Build and maintain a model benchmarking framework with ground truth datasets, automated precision/recall/F1 measurement, and AUC-ROC analysis.
• Implement threshold tuning workflows to calibrate alert sensitivity per desk and jurisdiction, ensuring compliance with ASIC INFO 283 guidance.
• Develop false positive and false negative reduction strategies, including contextual filtering, trader baseline profiling, and red team simulated misconduct testing.
• Deliver model explainability using SHAP/LIME for regulatory audits and produce monthly effectiveness scorecards for the compliance committee.

Required Skills

• 5+ years in machine learning engineering, with at least 3 years in NLP or text classification within financial services or compliance.
• Strong proficiency in Python and libraries including scikit-learn, PyTorch, TensorFlow, and Hugging Face Transformers.
• Hands-on experience fine-tuning pre-trained language models (BERT, RoBERTa, GPT-family) for domain-specific tasks.
• Experience building model evaluation pipelines with automated metric tracking, drift detection (PSI, KL divergence), and champion-challenger evaluation.
• Understanding of financial services compliance, including market abuse, insider trading, and conduct risk.
• Familiarity with model explainability frameworks (SHAP, LIME) and model governance requirements.
• Bachelor’s or Master’s degree in Computer Science, Machine Learning, Statistics, or Computational Linguistics.

Key Skills

Education

Bachelor's or Master's degrees