Computational Scientist II

• Build and Deploy Agentic LLM Workflows
  • Design and implement LLM agent-based pipelines that automate or augment complex scientific workflows within the CPP group.
  • Develop human-in-the-loop systems that allow scientists to collaborate with AI tools through natural language, iterative feedback, and structured outputs.
  • Integrate domain-specific context — such as internal guidelines, templates, and scientific reference materials — into LLM workflows to ensure outputs meet scientific and regulatory standards.
  • Package reusable LLM workflow components and libraries, and ensure tools are production-ready, well-documented, and accessible to scientist users with varying technical backgrounds.
• Develop Quality and Evaluation Infrastructure
  • Build automated quality control layers to evaluate LLM outputs against structural, scientific, and consistency criteria.
  • Design evaluation frameworks to measure output quality, efficiency gains, and failure modes over time.
  • Maintain versioned evaluation logs and contribute to periodic reports supporting tool improvement and stakeholder communication.
• Collaborate and Innovate
  • Work closely with pharmacometricians, data scientists, and automation engineers to understand scientific requirements and translate them into robust system designs.
  • Stay current with advances in LLM tooling, agentic frameworks, and AI applications in drug development and R&D.
  • Contribute to the broader DevSci AI adoption by sharing learnings and best practices across functions as well as providing trainings.

• You hold or are pursuing a Master's or PhD degree in a quantitative or computational field, such as Computer Science, Data Science, Bioinformatics, Computational Biology, Pharmaceutical Sciences, or a related discipline.OR BA/BS w/ 5 years min exp
• Familiarity with the drug development or pharmaceutical R&D context is strongly preferred. Candidates with an awareness of clinical development processes, quantitative sciences, or life sciences research — and a genuine interest in applying AI to advance drug development — will thrive in this role.
• Software and AI/LLM Engineering
  • Solid software engineering experience and familiarity with standard development practices — version control (Git), code review, documentation, and working effectively within collaborative codebases are expected. Comfortable working with structured and unstructured data, and able to pick up new tools and frameworks quickly. Prior exposure to LLM applications is a big plus — whether through building and deploying LLM-powered pipelines, working with agentic frameworks such as LangSmith, retrieval-augmented generation (RAG) architectures, harness engineering or guardrail design, or integrating LLM APIs into functional tools. Familiarity with cloud platforms (AWS, GCP, or Azure) and a portfolio of relevant projects are also welcomed.
• Collaborative and communicative — able to bridge engineering and scientific perspectives and explain technical decisions clearly.
• Self-directed and ownership-oriented, with a track record of delivering in ambiguous or evolving project environments.
• Curious and motivated by the intersection of AI technology and real-world scientific impact.

#LI-AS1