Scientist Manager, Bio-AI & ML Ops

About the role

P&G is looking for a Scientist Manager to lead the design, development, evaluation, deployment and scaling of predictive Bio-AI and integrated Consumer data models and multi-agent workflows. The role owns turning trusted data, scientific knowledge and domain tools into governed AI solutions that predict outcomes, generate and evaluate hypotheses, guide experimentation and accelerate innovation. It sits at the P&G Singapore Innovation Center at Biopolis, reports to a local R&D Director and joins a growing team of functional and technical experts.

Key responsibilities

• Build predictive and agentic AI solutions: multi-agent scientific workflows for evidence retrieval, hypothesis generation and evaluation, mechanistic reasoning, experiment planning and decision support
• Build RAG and Graph-RAG solutions across internal and external sources such as biological databases, scientific literature, patents, ChEMBL, supplier data and formulation knowledge
• Enable agents to use internal databases, APIs, simulation tools, OMICS pipelines, dashboards and scientific models, and partner with scientists, bioinformaticians and data engineers to turn workflows into agent-ready tasks
• Evaluate emerging methods in agentic AI, AI Scientist systems, biological foundation models and AI Co-Scientist architectures
• Set up evaluation frameworks covering predictive performance, task success, evidence quality, provenance, hallucination risk, reasoning quality, reproducibility and human-review outcomes
• Implement production controls for model, agent, prompt and workflow versioning, observability, tracing, CI/CD, access control, monitoring and auditability
• Turn pilots into reusable platform capabilities, scale validated workflows and build closed learning loops from experimental outcomes and scientist feedback

Required profile

• Bachelor's, Master's or PhD in Data Science, Bioinformatics, Computational Biology, Biostatistics or a related field
• Experience developing and deploying predictive models, LLM applications, retrieval systems or agentic workflows
• Experience designing evaluation approaches for machine-learning, LLM or agentic AI systems

Required skills

• Strong Python engineering and hands-on experience delivering AI, machine-learning or data solutions
• Tool calling, structured outputs, workflow orchestration and agent state or memory
• Familiarity with APIs, cloud deployment, containers, CI/CD, monitoring and ML Ops or LLM Ops
• Preferred: RAG, Graph-RAG, vector databases, tool-use architectures, agent orchestration, biological or consumer data, scientific knowledge graphs, ontologies, pathway databases, literature and patent mining, ELN/LIMS integration, human-in-the-loop systems, biological foundation models, multimodal scientific AI, active learning or AI Co-Scientist systems