Principal Data Scientist

Company: Marshmallow
Apply for the Principal Data Scientist
Location: London
Job Description:

Overview

As Principal Data Scientist in Claims, you lead data science initiatives across fraud risk and automated claims, shaping scalable ML and Generative AI solutions. You will collaborate with Product and Engineering to align technical direction with business goals and drive end-to-end decisioning improvements. You’ll champion robust, measurable impact through production-ready systems and clear communication with senior stakeholders. This role offers the chance to influence multi-year AI-powered claims automation and platform investments with a real-world impact.

Pay / Benefits

  • Flexible working: hybrid London office
  • Competitive bonus scheme
  • Flexible benefits budget (Ben Mastercard)
  • Mental wellbeing support via Oliva
  • Learning and development budgets and 2 extra training days
  • Private health care and Vitality benefits

Responsibilities

  • Provide technical leadership for data science across Claims Fraud and risk decisioning
  • Design, build and iterate production ML and Generative AI/LLM systems for claims validation and automation
  • Collaborate with Claims data scientists to align models, data, and workflows at scale
  • Advocate for platform and tooling investments (monitoring, feedback loops, QA) for end-to-end automation
  • Influence technical direction with Product and Engineering to support long-term automated claims handling
  • Set high standards for statistical rigour, experimentation, and measurement for stakeholder communication

Key requirements

  • Significant commercial experience delivering end-to-end ML solutions
  • Hands-on experience building and shipping Generative AI systems in production
  • Strong statistical and modelling foundation with risk-based decisioning experience
  • Proven ability to influence technical direction across Data Science and Engineering
  • Strong stakeholder management and ability to push back constructively with cross-functional teams
  • Cross-functional collaboration
  • Effective communication across seniority levels
  • Adaptability in ambiguity and change
  • Production ML systems
  • Generative AI / LLM in production
  • Risk-based decisioning under uncertainty

Posted: September 14th, 2026