Senior Data Scientist, Payments Foundation Models

Visa Visa · Fintech · Cambridge, United Kingdom, United Kingdom

Visa is building a new Payments Foundation Models team to develop next-generation AI models for fraud and identity solutions, with potential applications in credit risk and personalized commerce. The Senior Data Scientist will train, evaluate, and document these models, collaborating with cross-functional teams to improve their quality and adoption. This role involves hands-on development, end-to-end data processing, deep learning model training (self-supervised, fine-tuning), and ensuring models navigate risk management processes. The position also includes leading deployment, collaborating with data engineers, communicating insights, and mentoring junior team members.

What you'd actually do

  1. Developing, training, evaluating, documenting and disseminating Payments Foundation Models for use in data science and AI projects across Visa.
  2. Collaborating across the organization with engineering, data science, research, product and commercial teams to improve the quality, adoption and real-world impact of our models.
  3. End-to-end processing and modelling of large data sets
  4. Training deep learning models utilizing self-supervised training, supervised fine-tuning or adaptation approaches.
  5. Ensuring new deep learning models successfully navigate model risk management processes, ensuring high quality documentation exists alongside analytics products (reports, presentations, visualizations)

Skills

Required

  • Python
  • SQL
  • PyTorch
  • deep learning framework
  • artificial neural networks
  • model risk management framework

Nice to have

  • R
  • big data tools
  • git
  • fraud detection
  • risk analytics
  • financial crime prevention
  • banking and payments industry

What the JD emphasized

  • Payments Foundation Models
  • foundation AI models
  • deep learning models
  • model risk management processes

Other signals

  • building foundation AI models
  • payments-focused foundation AI models
  • fraud scores
  • credit risk modelling
  • agentic commerce personalization