Senior Data Scientist - RDT ERP (Full-Stack AI Developer)

About the role

Roche's D2P Data & Analytics area leads end-to-end Procurement, Accounts Payable and Travel & Expense Management data and AI strategies, turning functional strategies into measurable business outcomes. It is looking for a hands-on Full-Stack AI Developer (Generalist) to join a small, fast-moving team and work with global platforms. The role suits someone who enjoys building end-to-end products, thrives in ambiguity and wants to apply AI in practical, user-facing applications. The position involves collaboration with Global Procurement, Diagnostics Direct Procurement and RSS Operations.

Key responsibilities

• Design, build and ship AI-powered applications from concept to production
• Develop backend services and APIs using Python
• Build responsive frontend interfaces using JavaScript / TypeScript
• Integrate LLMs into products for real-world use cases
• Build and improve RAG pipelines, prompt workflows and agentic AI systems
• Use modern AI coding tools to speed up development and delivery
• Work closely with product and business stakeholders in a fast-paced environment
• Own features across architecture, implementation, testing and deployment

Required profile

• Bachelor's degree in Computer Science, Software Engineering or a related field preferred; equivalent practical experience is fully acceptable
• Demonstrated experience independently identifying analytical opportunities, exploring data deeply and owning projects from start to deployment
• Proven track record of driving medium-sized projects that influence business decisions
• A strong portfolio or GitHub profile with personal projects is highly valued
• Exceptional communication skills, able to explain complex findings to technical and non-technical audiences
• Product-minded, pragmatic, fast learner, self-driven and team player

Required skills

• Python backend development and JavaScript / TypeScript frontend development
• Full-stack capability, shipping complete features
• AI / GenAI: prompt engineering, RAG pipelines, LLM integrations and AI agents
• Frameworks such as LangChain, LangGraph or similar AI orchestration tools
• Active use of AI-assisted coding tools (e.g. Claude Code, Cursor, GitHub Copilot) as part of daily work
• Snowflake Cortex AI or Dataiku (optional)