Job Description
The Fraud team at Wise is dedicated to safeguarding our platform against financial crime and ensuring the protection of our legitimate customers. Leveraging cutting‑edge machine learning, real‑time transaction monitoring, and data analysis, our team is responsible for developing and enhancing fraud detection systems. Software engineers, data analysts, and data scientists collaborate on a daily basis to continuously improve our systems and provide support to our fraud investigation team.
Vision
* Build a globally scalable fraud prevention and detection engine to maintain Wise as a secure environment for our legitimate customers.
* Utilise machine learning techniques to identify potential risks associated with customer activity.
* Foster a strong partnership between our fraud investigators and the product team to develop solutions that leverage the expertise of fraud prevention specialists.
* Not only meet the requirements set by regulators and auditors but also surpass their expectations.
Responsibilities
* Innovate and Develop: Lead the development and deployment of machine learning models, including neural networks, anomaly detection, graph‑based models, Transformer‑based models.
* Lead and Collaborate: Mentor team members and promote adoption of AI workflows for automation across the business. Collaborate with cross‑functional teams to integrate data science solutions into fraud‑prevention product offerings.
* Deploy and Integrate: Develop scalable deployment strategies together with Platform teams and integrate LLMs with AI agents for seamless production use.
* Optimise and Evaluate: Conduct large‑scale training and hyper‑parameter tuning, and define performance metrics to ensure high‑quality model outputs.
* Data Strategy and Management: Design and implement strategies for data collection, curation, and augmentation to support robust model training.
* Documentation and Reporting: Communicate complex data findings to non‑technical stakeholders effectively. Document the development and maintenance processes for models and features.
#J-18808-Ljbffr