SlipstreamJobs tracks this role from the company's public career site. Apply directly on the employer's site.
Marshmallow is a unicorn fintech company (founded 2017, £140M+ raised, 700+ employees) on a mission to make financial services accessible to underserved populations, particularly migrants new to the UK. The company has built insurance and car finance products and is expanding rapidly toward becoming a major global financial services provider.
The Car Finance team is building a new lending platform from the ground up, designed to serve newcomers to the UK with fairly priced car finance. The team combines hands-on operations with automation and AI-powered processes to move faster than traditional lenders.
In this role, you'll sit at the core of credit decisioning: determining who receives loans, at what price, and how to measure success. Marshmallow's customers are people traditional lenders struggle to assess—many are new to the UK with thin or no credit files. The challenge is to combine bureau data, Open Banking, insurance, application, and behavioral data to build a risk view that is fair, explainable, and compliant.
You'll work between a traditional credit analyst and data scientist, owning work end-to-end: framing questions, pulling data, modeling, presenting recommendations, and seeing decisions live in the decision engine. You'll build and monitor scorecards, design and test credit strategies, dig into portfolio performance, and turn messy datasets into actionable insights. You'll work with a small, high-performing team and a Head of Credit who has built the function from scratch, giving you real ownership from day one. This is not a handoff role—you'll see the impact of your decisions immediately.
Key responsibilities:
- Build, validate, and monitor application and behavioral scorecards using bureau, Open Banking, and internal data to assess new-to-UK and thin-file customers.
- Design, test, and refine credit strategy: cut-offs, affordability rules, pricing tiers, and limits, balancing risk, volume, and margin.
- Run experiments and champion/challenge tests to validate changes before scaling.
- Explore new data sources and alternative signals; build the case for using (or not using) them in decisioning.
- Support model governance: document work, evidence fair treatment, and meet Consumer Duty, CONC, and model risk obligations.
- Work with engineers to deploy models and rules into production; use AI tooling to accelerate team velocity.
Requirements:
Must have:
- 2–5 years' experience in decision science or data science within consumer lending or regulated financial services.
- Hands-on experience developing or materially refining credit scorecards or risk models (logistic regression and/or gradient boosting), including sample design, binning, reject inference, validation, and monitoring.
- Experience working with Open Banking or other alternative data in credit decisioning.
- SQL proficiency and comfort with large, imperfect datasets.
- Solid grounding in statistics and good instinct for distinguishing real results from noise.
- Commercial judgment: ability to translate model output into lending decisions and explain trade-offs to non-technical audiences.
- Curiosity and ownership: you ask why, dig for answers, and prefer shipping and improving over waiting for perfection.
Nice to have:
- Motor finance or other secured lending experience.
- Experience in rapidly scaling lending environments.
- Familiarity with UK consumer credit regulation and its impact on model design and use.
- Experience using LLMs to accelerate analytical work.
- Familiarity with Marshmallow's stack: Snowflake, Python, Claude, Linear, Notion.
