Data Scientist - Talent Pool ( Sandton) at Discovery Limited

Key Purpose
The Data Science Lab is a highly specialised and expanding team that tackles challenges in the health, life, and short-term insurance businesses, as well projects that cut across the whole of the Discovery Group. We are looking for individuals, across all levels and years of experience, for new projects related to:

• risk management through behavioural science and intervention (next best action) design 
• natural language processing & large language models
• combining traditional data (eg: wearable device, web & app) with novel data sources in new ways
• driving improvements to customer and advisor experience metrics
• optimising process efficiencies

Areas of responsibility may include but not limited to

• Identify and build appropriate models to predict risk, sales and savings
• Present data and model findings in a way that provides actionable insights for business users
• Mining and visualising large structured and unstructured datasets throughout the businesses to find new insights to inform product design, risk management, customer interaction strategies, etc.
• Following model implementations through to business adoption
• Monitoring model performance and using feedback for improvement
• Improving processes and data collections where opportunities arise
• Running scientific experiments to evaluate different models in a reproducible way

Personal Attributes and Technical Skills  

• A creative and enthusiastic attitude to unearthing valuable insights and generating value for Discovery clients 
• Ability to balance multiple priorities and to step back and see how analytics work fits into the wider business context 
• Aligned to Discovery values and core purpose 

Education and Experience

• Master’s or PhD degree in either Data Science, Actuarial Science, Statistics, Operations Research or Applied Mathematics with some experience in software engineering, computer science or working with big disparate sets of data, OR
• Master’s or PhD degree in Computer Science or Software Engineering with solid experience in statistical modelling, data mining, machine learning or optimisation
• Other analytical qualifications will also be considered if accompanied by the relevant experience
• Good experience with writing queries in SQL 
• Some exposure to big data tools and programming languages is preferred, e.g. Spark, Scala 
• Ability to formulate a clear problem statement, develop a plan for tackling it, and clearly communicate findings verbally, visually, and in writing 
• Ability to source and combine data from both structured and unstructured sources 

Advantageous:

• Using and/or developing NLP packages and models
• Using machine learning frameworks (e.g. PyTorch, TensorFlow) for training deep learning models
• Using adapting, and fine-tuning Large Language Models (LLMs)