Senior Quantitative Analytics Specialist

Advanced Model Validation (Primary Focus)

• Technical Auditing: Perform end-to-end validation of high-complexity AI/ML models, including Deep Learning architectures (CNNs, RNNs, Transformers).

• Conceptual Soundness: Evaluate the mathematical and logical foundations of fraud models to ensure they are fit for purpose and resilient to adversarial attacks.

• Performance Testing: Conduct rigorous back-testing, sensitivity analysis, and stress testing of models to identify potential biases or failure points in high-risk fraud scenarios.

• Risk Assessment: Quantify Fraud Risk and model limitations, providing clear "Go/No-Go" recommendations to senior leadership and regulatory bodies.

2. Multi-Modal Fraud Expertise

• Voice Fraud: Validate models designed to detect synthetic voice, voice cloning, and social engineering in audio streams.

• Image Fraud: Challenge computer vision models used for document verification, facial recognition, and the detection of digital image manipulation (Deepfakes).

• Pattern Recognition: Evaluate ML models that identify behavioral fraud patterns and anomalies in transactional data.

3. Development & Lifecycle Support

• Collaboration: Partner with Model Development teams during the design phase to ensure "Validation-ready" architecture.

• MLOps Integration: Monitor model performance in production, ensuring that model decay or data drift is detected and mitigated via automated feedback loops.

• Compliance: Ensure all models adhere to internal model risk management (MRM) policies and external regulatory requirements.