Data Scientist, (ai/ml) – Visa Consulting and Analytics

Visa · Fintech · Toronto, Canada, CA

Visa is seeking a Data Scientist with AI/ML experience to join their Global Data Science group. The role involves building predictive models, applying GenAI techniques to transaction data, and developing agentic AI systems for payment decisioning. The position requires experience with LLM orchestration frameworks, vector databases, and embedding models, and will leverage AI coding tools to accelerate development. The role is focused on applying AI/ML to innovate and optimize customer experiences, revenue generation, and business outcomes within the fintech domain.

What you'd actually do

  1. Build agentic AI systems with multi-step reasoning, tool use, and memory for complex payment decisioning workflows
  2. Apply data science and GenAI techniques to analyze transaction data and deliver actionable client insights
  3. Leverage AI coding tools (e.g., GitHub Copilot, Claude Code, Cline, OpenAI Codex) to accelerate development
  4. Use and build new predictive models to innovate and optimize customer experiences, revenue generation, data insights, advertising targeting and other business outcomes
  5. Work with partners throughout the organization to explore opportunities for using Visa data to drive business solutions

Skills

Required

  • Python
  • SQL
  • LLM orchestration frameworks
  • vector databases
  • embedding models

Nice to have

  • financial services
  • credit cards
  • merchant analytics

What the JD emphasized

  • agentic AI systems with multi-step reasoning, tool use, and memory for complex payment decisioning workflows
  • LLM orchestration frameworks (LangChain or similar), vector databases, and embedding models

Other signals

  • Build agentic AI systems with multi-step reasoning, tool use, and memory for complex payment decisioning workflows
  • Apply data science and GenAI techniques to analyze transaction data and deliver actionable client insights
  • Leverage AI coding tools (e.g., GitHub Copilot, Claude Code, Cline, OpenAI Codex) to accelerate development