Senior Scientist, AI Applications – R&D Immunology Data Science & Digital Health Full-time

Senior Scientist, AI Applications

Johnson & Johnson Innovative Medicine is currently seeking a Senior Scientist, AI Applications to join our Immunology R&D Data Science & Digital Health team (DSDH). This position will be located on site at one of our offices in either Spring House PA, Cambridge MA, Titusville NJ, Raritan NJ, or San Diego CA (NO fully remote option available).

The R&D Immunology Data Science and Digital Health (DSDH) team is seeking a Senior Scientist, AI Applications – Immunology to work closely with the Machine Intelligence (MI) team within DSDH AI/ML to develop and deliver artificial intelligence (AI)/machine learning (ML) scientific applications. This must be a mission-driven Scientist aimed at developing the future of autonomous AI in Immunology drug discovery and development. You'll work with Large Language Models (LLMs) and sophisticated data science techniques, harnessing massive datasets spanning scientific literature, omics, and biomolecular knowledge. Your work will directly accelerate research across the drug development pipeline, including novel target discovery, translational research and clinical trials—making reasoning agents faster, more accurate, and more reliable. Key focus areas include multimodal reasoning, agent robustness and uncertainty, agent memory architecture, and agent motivation frameworks. This role will transform AI/ML needs into actionable science, and work hands on to deliver decision-driving results.

You will connect research across the entire R&D lifecycle so broad knowledge of how advanced analytics is applied across Johnson and Johnson Innovative Medicine pipeline is expected.

Key Responsibilities:

• An individual contributor position focused on hands-on technical delivery of advanced analytic (AI/ML) Immunology R&D research.
• Work with immunology R&D and Data Science business leaders to translate business opportunities into hands on research plans.
• Develop and optimize multimodal reasoning systems that integrate large-scale structured and unstructured data to derive actionable conclusions
• Build novel agent memory architectures to enable persistent knowledge retention and context-aware decision-making across extended discovery workflows
• Prototype and evaluate innovative approaches for uncertainty quantification and reliability assessment in AI-driven scientific reasoning
• Serve as a domain expert within the field by articulating to team members emerging approaches and methodologies within the AI/ML with a clear understanding of their practical applications within the industry
• Benchmark agent performance against human expert decisions and existing computational methods and demonstrate ability to publish results in leading journals and conferences
• Define and enforce engineering best practices across AI prototyping, including architecture patterns, documentation standards, quality gates, and reproducibility criteria
• Serve as a key contributor to the over-arching AI strategy and to enable the scaling and deployment of tools aimed at the broader R&D community.

Required qualifications:

• A Ph.D. degree in AI/ML or related field (e.g., computer science, machine learning, data science, applied mathematics, statistics) with at least two (2) years working experience
• Knowledge in sophisticated mathematics areas such as Bayesian inference, uncertainty quantification, information theory, causal inferencing, graph neural networks, etc.
• Proficiency with open-source agentic frameworks (e.g., LangGraph, DSPy, mem0, etc.) and deep learning frameworks (e.g., PyTorch, Tensorflow, etc.)
• Proficiency in one or more programming languages – preferably Python
• Positive and motivated demeanor with the ability to work both independently and in a diverse, multi-disciplinary team

Preferred qualifications:

• Understanding of the drug discovery pipeline and familiarity with biological data types
• Experience in Immunology research
• Strong understanding of statistics and applicable methods to validate model performance
• Familiarity with cloud hosting solutions (ex. AWS and/or Azure)