Amgen harnesses the best of biology and technology to fight the world’s toughest diseases, making people’s lives easier, fuller, and longer. We discover, develop, manufacture, and deliver innovative medicines to help millions of patients. Amgen helped establish the biotechnology industry more than 40 years ago and remains on the cutting edge of innovation, using technology and human genetic data to push beyond what’s known today.
The Data Scientist – Agentic AI & Scientific Systems is a senior technical contributor responsible for designing, building, and integrating AI capabilities that accelerate scientific discovery across domains such as protein engineering, structure prediction, disease biology, and target identification.
This role focuses on developing agentic AI systems and scientific AI workflows that combine foundation models, domain-specific models, knowledge sources, and computational tools into reusable solutions that support scientific decision-making.
The engineer works closely with scientific domain leads to translate research needs into scalable AI solutions and reusable capabilities. This role serves as a bridge between scientific innovation and enterprise AI platforms, helping establish a foundation for next-generation AI-assisted scientific workflows.
Agentic AI Systems Development
Design and implement agent-based systems that support complex scientific workflows.
Develop capabilities including:
Build reusable components for:
Evaluate emerging agent frameworks and contribute to standards and best practices across projects.
Scientific AI & Model Integration
Integrate foundation models and scientific AI models into end-to-end workflows.
Develop reusable APIs, services, and interfaces that allow AI agents and applications to consume scientific models and computational tools.
Collaborate with scientific domain experts to identify appropriate modeling approaches and evaluate solution effectiveness.
Knowledge Systems & Retrieval
Design and implement knowledge-driven AI systems that connect LLMs and agents with enterprise and scientific data.
Develop solutions utilizing:
Ensure AI systems leverage authoritative knowledge sources and support traceability and explainability.
Develop end-to-end workflows that combine:
Create reusable workflow patterns that can be applied across multiple scientific domains and projects.
Contribute to architectural decisions regarding workflow design, model integration, and AI system composition.
Evaluation & Responsible AI
Develop evaluation frameworks for AI systems, agents, and workflows.
Establish approaches for measuring:
Support responsible AI practices including transparency, traceability, and governance requirements.
Collaboration & Scientific Partnership
Translate scientific requirements into technical solutions and provide guidance on AI capabilities, limitations, and implementation approaches.
Contribute to technical design reviews and mentor junior team members where appropriate.
Strong engineering background in AI and machine learning systems.
Hands-on experience with:
Agent frameworks (LangGraph, LangChain, AutoGen, CrewAI, Semantic Kernel, or similar)
Ready to Apply for the Job?
We highly recommend utilizing Workday's robust Career Profile feature to complete the application process. A link to update your profile is available when you click Apply. You can then complete your Workday profile in minutes with the “Upload My Experience” functionality to upload an updated copy of your resume or you can simply edit the individual sections of your Career Profile.
Please note that you should be in your current position for at least 18 months before applying to internal positions. Staff must notify their current manager if invited for an interview. In addition, Staff are ineligible to apply for open positions if (a) their performance is currently being managed on a performance improvement plan (PIP) or other locally utilized formal coaching document or (b) their most recent performance rating was not a “Partially Meets Expectations” or higher. Please visit ourInternal Transfer Guidelines for more detailed information
