Health Data Scientist (Gen AI & Agentic AI)

About the role

The Health Data Science Team in Singapore works with global teams to deliver machine learning, advanced analytics, disease modeling and data asset and infrastructure capabilities. It builds ML models from multimodal data (structured data, free text, images, video and audio) to power digital health solutions and maintains governed data platforms and pipelines. The team is looking for Health Data Scientists to contribute to projects across health and technology domains.

Key responsibilities

• Collaborate with internal and external stakeholders globally on digital health solutions, real-world evidence and clinical data initiatives.
• Lead the development of GenAI and multimodal, multi-agent solutions using text, image, audio and video capabilities.
• Design interactive analytical frameworks for the governed analysis of real-world and clinical data.
• Establish testing and monitoring processes to ensure the accuracy, reliability and performance of generative and agentic systems.
• Communicate insights, outcomes and recommendations to stakeholders within Danone and external partners.

Required profile

• Master's or Ph.D. in a relevant technical discipline.
• At least 5 years of experience in data science within healthcare, life sciences or digital health.
• Strong stakeholder management and collaboration across cross-functional and global teams.
• Ability to turn business challenges into data science and AI initiatives and to explain complex technical concepts to non-technical audiences.

Required skills

• Deep learning expertise and at least one leading deep learning framework.
• MLOps pipelines and end-to-end machine learning lifecycles.
• Multimodal generative AI: VAEs, GANs, diffusion models, ViT, Swin Transformers, LLMs (BERT, GPT), fine-tuning, LoRA, knowledge distillation, vision-language models.
• Multi-agent systems: memory architectures, vector databases, RAG, reasoning and planning agents, tool integration, MCP, agent orchestration, guardrails, reinforcement learning, automated evaluation.
• Experience with Agent Bricks and Strands Agents is advantageous.
• AWS and/or Azure; Agile (Scrum, Kanban) highly desirable.