Senior Data Scientist

The team you will be joining: NorthAm Chargeback Team

Our North America Chargeback team is responsible for mitigating chargeback risk for popular US payment systems such as ACH (Automated Clearing House).

The North America Chargeback team uses sophisticated machine learning models to predict the chargeback risk for ACH payments.

We are looking for a capable and experienced Senior Data Scientist to be responsible for machine learning models for chargeback risk mitigation. You will own this function - be responsible for the development and maintenance of existing and new models. Together with product and engineering teams, you will develop a vision for effective chargeback risk mitigation of US payment methods.

Here’s how you’ll be contributing:

• You will help improve operational-risk decisions, digging into huge volume of data to find insights to identify and validate new opportunities, and then bring these to life via production grade models that can be clearly monitored by our operations team

• You will introduce state of the art modelling to the domain, levelling-up the existing models to harness the latest deep learning developments in the field of fraud and suspicious activity detection

• You will help the data science team develop models for anomaly detection through prototyping model features and developing them into production ready pipelines

• You will be using our tech stack (Airflow, AWS Sagemaker, Flink, Spark, etc..) to build machine-learning workflows for automatic model training, testing, monitoring and deployment

• You will design, execute and evaluate experiments to automate operational processes using Large Language Models and other data science technologies

• Your average day will include building new models, maintaining production models, evaluating new ideas, communicating findings with your team, putting out fires etc.

What does it take?

• 4+ years experience in building machine learning models on large datasets, using the right tools depending on the data volumes (we use Python, Spark, SQL, etc.)
• A solid knowledge of Python, and are able to make and justify design decisions in your code. You know how to use Git to collaborate with others (e.g. opening Pull Requests on GitHub) and are able to review code
• Familiarity with a range of model types, and know when and why to use gradient boosting, neural networks, regression, autoencoders, clustering or a blend of these
• An excellent understanding of statistics and can use experiments to derive decisions with degrees of certainty
• Confidence in managing your own workload in order to deliver projects in a self-supervised manner, while effectively communicating your progress to your team and the wider organisation

Some nice-to-haves (but not essential):

• Previous experience in fraud detection and / or anomaly detection
• Work experience in financial institutions and knowledge of payment systems
• You have a Mathematics / Exact sciences / Engineering / Finance background
• Understanding of Object-oriented programming and ability to read Java code