Senior Fraud Data Scientist

PayInc is a purpose-driven payments provider building on over 50 years of trusted history in South Africas payments ecosystem. Our mission is to connect people businesses and economies through secure efficient and inclusive digital payments infrastructure and be a catalyst for financial inclusion and economic growth. From EFTs and cards to PayShap PayInc provides the backbone that enables money to move safely across the economy. At our core we exist to make great connections empowering participation enabling growth and ensuring no one is left behind.

PURPOSE

As a Senior Fraud Data Scientist you play a key role in supporting the delivery of the Fraud Intelligence service by leveraging data to discover new insights and develop solutions that allow for improved decision making. This role requires strong quantitative technical and analytical skills to balance two competing demands - enabling a frictionless customer experience while minimising fraud risk and money laundering on real time electronic payment platforms. You will be required to use statistical and machine learning techniques to maximise the performance of systems and design algorithms and heuristics to identify high risk transactions automating real time transaction decisions. You are expected to apply and leverage off toolkits multiple skillsets data engineering advanced computing scientific methods statistical computation visualization business communication domain expertise and associated data to contribute directly to the proactive detection of fraud and the reduction of fraud losses while developing optimizing maintaining and evolving fraud detection and performance models at a national level that is imperative to improving scoring performance and account for shifting fraud patterns.

You will engage with the following stakeholders:
Fraud Team (Business Owner Analytics and Detection Team Stakeholder Relationships)
Internal departments (IT Ops: Infrastructure Networks Applications Database Service Desk)
Service providers industry bodies & vendors

Your key responsibilities include:

Develop deploy and maintain fraud detection rules and scoring models across PayShap RTC EFT and ACD payment rails
Design and build machine learning models for fraud scoring incorporating entity state profiling behavioural analytics and network analysis techniques
Conduct fraud universe coverage analysis by combining system performance metrics with confirmed fraud data identifying detection gaps and prioritising rule/model enhancements
Work with cross-departmental teams to define metrics guidelines and strategies for effective use of algorithms and data
Establish and maintain coding standards statistical reporting methodologies and data analysis best practices
Coordinate data resource requirements between analytics and technical teams
Work with product managers engineers and analytics team members to translate prototypes into production
Identify fraud patterns through the monitoring and analysis of transactions across all payment streams
Prepare and deliver client-facing presentations and reports explaining fraud detection performance scoring mechanisms and rule behaviour to participant banks
Conduct research and make recommendations on data infrastructure database technologies analytics tools services protocols and standards
Drive the collection of new data and the refinement of existing data sources
Develop algorithms and predictive models to reduce the frequency of fraudulent transactions
Develop tools and fraudulent transaction libraries that will help analytics team members more efficiently flag fraudulent transactions
Contribute to the development and assessment of alternative fraud detection capabilities including potential platform replacement strategies
Mentor and support junior team members contributing to skills transfer and team development

QUALIFICATIONS / KNOWLEDGE
Minimum Qualification:
Degree (Honours Masters or PHD) in Statistics Computer Science Engineering Mathematics or a combination of these

Technical Knowledge
Machine learning techniques and frameworks (scikit-learn Tensorflow Pytorch or similar)

Python Programming (panda NumPy matplotlib seaborn) for data analysis and model developmentPySpark for large scale data processing (AWS Glue jobs)
SQL proficiency particularly with cloud based data warehouse (AWS Redshift preferred)
Cloud infrastructure experience (AWS services S3 Glue Sagemaker Redshift)
Data analytics life cycle and data engineering
Prescriptive and Statistical modelling
Fraud detection models and real time scoring systems
Data wrangling and feature engineering
Version Control (Git) and collaborative development practices
Familiarity with real time event processing and fraud detection platforms is advantageous
Expert in MS Office

Desirable Certifications
AWS Certified (Cloud Practioner Solutions Architect or Machine Learning Speciality)
Any recognised data science or machine learning certification (Courses Udemy Datacamp)

EXPERIENCE
5 years data science experience preferably in the financial services or payments industry
Experience deploying machine learning models to production environments Experience with big data analytics and large scale transaction datasets
Experience in fraud analysis risk analysis and payment risk management Experience in financial industry focusing on payment fraud
Good written and oral communication skills
Good interpersonal skills (require a patient and empathetic attitude)
Have strong time management and organisational skills (must be able to organise and manage multiple tasks at a time)
Comfortable working in fast paced environment
Ability to work autonomously and in teams
Good troubleshooting and problem solving skills