The Machine Learning group within the CCB Risk Fraud Modeling team is responsible for developing and implementing best-in-class fraud prevention and detection models and analytical tools. The team provides diverse models and analytical tools used to identify potentially fraudulent transactions across different lines of business (card, retail, auto, merchant services).
Working for one of the largest banks, card issuers, and payments processors in the US, you will be fighting crime and protecting consumers and small businesses from financial fraud, including account takeovers and identity theft, with mathematical modeling. You will work in an industrial R&D/skunkworks environment, developing innovative predictive models on a dataset in the hundreds of TBs.
In this role, you will lead a small team of elite analytical experts to identify and retool suitable machine learning algorithms that can enhance the fraud risk ranking of particular transactions and/or applications for new products. This includes a balance of feature engineering, feature selection, and developing and training machine learning algorithms using cutting edge technology to extract predictive models/patterns from data gathered for billions of transactions. In addition, you will work with technology partners to develop and improve cutting-edge real-time and batch model execution environments that process hundreds of millions of transactions per day. You will coordinate the design, development, deployment and monitoring of production real-time machine learning systems that execute on many millions of transactions per day and impact up to half the households in the USA.
Masters degree in Mathematics, Statistics, Economics, Computer Science, Operations Research, Physics, and other related quantitative fields
At least 5 years experience with design and deployment of live production models implemented in Python with modern data science techniques such as neural networks or gradient-boosted trees
An expert who knows how models work, the reasons why particular models work or not work on particular problems, and the practical aspects of how new models are designed
Proven leadership in driving changes/delivering values and small team management experience
Proven ability to coordinate the work of a data science team with their technology, business and risk management partners
PhD in a quantitative field with publications in top journals, preferably in machine learning
Hand-on experience working with bare-metal hardware in a Linux shell
Experience with model design in a big data environment via Hadoop, Spark and Hive
Experience designing models with Keras/TensorFlow, PyTorch, or other frameworks on GPU hardware
Launch your career - Upload your resume now!Upload your resume
Loading some great jobs for you...