Applied AI Workflow Scientist, Sr. Manager

Pfizer Pfizer · Pharma · MA

This role focuses on applying LLMs and agentic AI to scientific workflows in healthcare (Pfizer). The scientist will identify use cases, build reusable workflows and agentic systems, and partner with scientists to ensure adoption and impact. The role emphasizes practical application, scientific credibility, and scalable solutions.

What you'd actually do

  1. Identify high-value, repeat scientific and analytical use cases where LLMs, agentic AI, and workflow automation can materially improve the speed, quality, consistency, or accessibility of work across I&I.
  2. Design, build, and refine reusable AI workflows, prompt/program structures, orchestration patterns, and agent-based tools that support end-to-end scientific narratives rather than isolated task completion.
  3. Partner closely with scientists, clinicians, computational biologists, and other stakeholders to understand real workflow pain points, define fit-for-purpose solutions, and iterate rapidly toward tools that are scientifically useful and operationally adopted.
  4. Translate emerging LLM and agent capabilities into practical scientific applications, balancing speed of experimentation with methodological rigor, grounded usage, human-in-the-loop design, and reusable implementation patterns.
  5. Contribute to technical standards for evaluation, documentation, guardrails, and workflow quality so that AIDE solutions are trusted, reproducible, and suitable for repeated use across teams and projects.

Skills

Required

  • Python
  • modern AI / ML tooling
  • collaboration and communication skills
  • methodological rigor
  • model limitations
  • evaluation
  • human oversight in AI-enabled workflows

Nice to have

  • life sciences
  • pharma
  • biotech
  • systems biology
  • immunology
  • translational science
  • omics
  • scientific evidence synthesis
  • literature and document workflows
  • retrieval-augmented approaches
  • multi-step knowledge workflows
  • agentic orchestration
  • prompt/program design
  • workflow automation
  • multimodal extensions of AI systems
  • domain fluency

What the JD emphasized

  • reusable AI workflows
  • agentic AI
  • workflow automation
  • scientific use cases
  • practical application
  • reusable capabilities
  • reusable implementation patterns
  • reusable AI workflows
  • agentic orchestration
  • multi-step knowledge workflows

Other signals

  • building reusable AI workflows
  • agentic AI
  • LLMs
  • scientific use cases
  • workflow automation