Principal Research AI Innovation Lead

About the role

We are seeking a Principal Research AI Innovation Lead to design, prototype, and scale AI-enabled capabilities that accelerate scientific research. This role will work across Research, AI, data, product, engineering, and enterprise technology teams to identify high-value opportunities, build practical LLM-enabled solutions, evaluate scientific quality, and create reusable capability patterns that improve how research teams use AI.

Responsibilities

• Partner with scientists and research leaders to identify high-impact opportunities where AI can improve research speed, quality, consistency, traceability, and decision-making.
• Help shape multi-year GenAI strategies, lead workstreams, and establish reusable building blocks - agentic frameworks, evaluation harnesses, retrieval and grounding components, tool servers, prompt and policy libraries, and provenance infrastructure - on which research programs build.
• Architect and personally implement the agentic system-of-systems that executes complex, long-horizon scientific workflows across research, including target evidence assembly, indication rationale construction, biomarker interpretation, translational synthesis, literature and evidence triangulation, and decision support, with explicit attention to inter-agent coordination, state and memory management, verification, recovery from intermediate failure, and lifecycle governance of agents in production.
• Establish the scientifically rigorous evaluation, benchmarking, and reliability standards that critical research AI systems expected to meet, including curated benchmark datasets, expert-reviewed reference standards, rubric-based assessments, hallucination and grounding metrics, calibration of uncertainty, longitudinal monitoring, regression gating in production, and the governance under which those standards are applied.
• Design mechanisms for incorporating expert feedback, scientific rationale, provenance, and research context into AI workflows so that systems and institutional knowledge improve over time.
• Collaborate with engineering, data, IT, security, legal, vendor, and platform teams to ensure prototypes are designed with appropriate governance, integration paths, and scalability in mind.
• Communicate AI opportunities, risks, limitations, evidence quality, and results clearly to scientific, technical, and executive audiences.
• Stay current with emerging AI methods, tools, vendors, and industry practices, and assess where they can create practical value for Research.

Qualifications

• Bachelor's Degree 8+ years of academic / industry experience
• Master's Degree 6+ years of academic / industry experience
• PhD 4+ years of academic / industry experience
• Advanced degree, such as MS, PhD, PharmD, or equivalent experience in a scientific, computational, or AI-related field.
• Direct experience in one or more research areas such as target identification and evaluation, indication expansion, drug repurposing, biomarker discovery, translational research, clinical evidence review, or portfolio decision support.
• Ability to interpret scientific evidence, assess analytical quality, and evaluate whether AI-generated scientific outputs are grounded, appropriately caveated, and defensible.
• Substantive hands-on architectural depth in modern AI and large language model methods, including agentic workflows, multi-agent orchestration, long-horizon task execution, GraphRAG, Model Context Protocol (MCP) auth patterns, and deep research workflows with reasoning models, with the depth to define reference architectures and architectural standards rather than to integrate vendor APIs.
• Experience building AI systems that reason across heterogeneous scientific evidence, including genetic associations, clinical outcomes, literature, omics data, assay data, real-world data, or other biomedical data sources.
• Experience designing AI evaluation frameworks, benchmark datasets, human-reviewed reference standards, rubric-based assessments, or scientific quality metrics.
• Experience developing or adapting deep learning models for biological, biomedical, or translational research applications, including fine-tuning biology-focused large language models, multimodal generative models, protein or sequence foundation models, representation-learning models, or other domain-specific AI systems.
• Experience working in innovation-lab, accelerator, startup, skunkworks, or rapid-prototyping environments.

Skills

• Hands-on AI product and prototyping experience
• Working fluency in drug discovery, translational science, or a related research domain
• Assess whether an AI output is scientifically sound, appropriately grounded, and useful for real research decisions
• Engage with scientists on topics such as target evidence, indication selection, biomarker interpretation, translational rationale, or clinical evidence
• Partnership with AI engineers to turn those needs into scalable systems

Benefits

Read more careers.bms.com/working-with-us.

Pay

Remote - United States - US $145,020 - $175,728

Schedule

US Exempt Employees flexible time off (unlimited, with manager approval, 11 paid national holidays (not applicable to employees in Phoenix, AZ, Puerto Rico or Rayzebio employees))

Phoenix, AZ, Puerto Rico and Rayzebio Exempt, Non-Exempt, Hourly Employees 160 hours annual paid vacation for new hires with manager approval, 11 national holidays, and 3 optional holidays