
Lead: Model Validation
- Stellenbosch, Western Cape
- Permanent
- Full-time
- Provide subject matter expertise, oversight and guidance with regards to the technical review and validation of models used in the organisation, particularly machine learning models.
- Ensure appropriate model governance and model risk management processes and standards are in place in order to manage model risk, improve and optimise model methodologies and usage.
- Lead and deliver independent validations across a range of models, including credit risk, operational optimisation, financial crime, and marketing campaigns.
- Provide technical oversight and mentorship to analysts, supporting skills development and professional growth.
- Design and implement challenger models using advanced machine learning methods.
- Research current and new machine learning algorithms to ensure proposed models are fit for purpose and market best practice is being followed.
- Ensure models meet governance, regulatory, and internal risk standards.
- Partner with model development teams and business stakeholders to provide effective challenge and assurance.
- Contribute to the ongoing enhancement of model validation methodologies, frameworks, and best practices.
- Have 6+ years of proven experience in model development in the field of Data Science and ideally direct model validation or review experience.
- Strong technical expertise in tree-based methods, neural networks, clustering, anomaly detection, and other advanced statistical or machine learning techniques.
- Exposure to cloud platforms (AWS preferred) and emerging technologies such as large language models is desirable.
- Demonstrated ability to mentor and guide junior analysts, identifying skills gaps and supporting development.
- Excellent communication and stakeholder engagement skills, with the ability to work effectively across multiple teams and disciplines.
- Able to design and deliver end-to-end on all but the most complex ML projects without supervision.
- Data mining for use in analyses and predictive modelling.
- Have had own model(s) reviewed and validated or directly validated models in the past.
- Presenting and defending findings to model developers and senior management.