Applied Data Scientist – Fraud Prevention

Job description

We are Creditspring, a new way of borrowing that focuses on its members and provides them with safe and efficient short-term financial products. We're a fast-growing FCA-regulated consumer credit company. We have members, not customers and we take a lot of pride in that! As one of the UK’s only subscription finance company in the market, we truly have a unique value proposition. Our mission is very clear; to improve the financial stability and resilience of our members. We do this through the products we provide, the partnerships we have, and our educational content. We want our members, and everyone in the UK to be able to better manage their finances and steer them away from high-cost, unregulated credit options. About the role We are currently looking for an experienced and detail-oriented applied data science and business analyst to join our Underwriting data science team with primary focus on fraud detection and mitigation. This is a mid-level applied or ‘full-stack’ data scientist role, ideal for someone with good command of the analytical and machine learning toolkit and desire to drive process and systems change based on the gained insights. You will be instrumental in shaping company’s fraud prevention initiatives using internal and external data, developing and implementing fraud detection models and providing monitoring and analytics in this area. This role will collaborate extensively with the colleagues from across the business (Data, Engineering, Underwriting, Operational Risk and Product teams), and is critical to support further platform growth and credit product innovation. Responsibilities Collect, process and analyse large and complex internal and external datasets to identify trends, risks and opportunities Design, develop and maintain fraud scoring, identity resolution and credit scoring machine learning models Interact with new and existing datasets and solutions providers to run retro analysis, A/B testing and POC exercises Review and test applicability of latest developments in fraud modelling to company’s operations (graph and network analytics, behavioural biometrics, real-time detection, adversarial thinking, AI agent networks and other techniques) Testing and integration of external API feeds into decisioning flow Monitoring, reporting and visualisation of insights and performance metrics Cross-team collaboration on incoming queries related to Fraud, AML and KYC verification cases What you'll need to succeed Prior experience in fraud prevention analytics, preferably within an SME or retail lending environment. Experience developing and deploying machine learning models in a local and cloud environment Strong command of statistical inference and supervised machine learning stack (scikit-learn, pandas, numpy, jupyter). Solid knowledge of Python for data extraction, transformation and analysis SQL proficiency for working with data from multiple sources including internal data and external feeds Demonstrated success in systems integration and analytics delivery Commercial awareness with strong communication skills and the ability to influence stakeholders Nice to have Lending, fintech and regulated sectors work experience Working with web applications, cloud data stacks and event driven architecture (we run on ruby on rails, python, aws, github) Hands-on working with credit bureau and open banking data. First-hand experience with decisioning SaaS platforms or AI agents Don’t meet all the listed requirements? Research shows that women and people of underrepresented groups often don't apply for jobs unless they're 100% qualified. As an equal opportunities employer, we know that diversity is a key part of our teams' successes - so if your experience doesn’t fit perfectly but this role excites you, we’d love for you to apply. We’re committed to Creditspring being an inclusive environment where employees feel welcomed, valued and listened to; we want you to thrive as your true self. Please note that the People Team is contactable only via . Unsolicited emails to other team members will not be actioned