Agentic AI Engineer: Rare Disease Drug Discovery in Berlin

Agentic AI Engineer: Rare Disease Drug Discovery

The opportunity

A fast-growing AI-driven biotech is building the next generation of drug discovery, focused on rare diseases. Most of the thousands of known rare conditions still have no approved therapy. This company brings together AI, large-scale biomedical data and deep pharmacology expertise to find new treatments faster and at lower cost than traditional R&D.

They’re now hiring an Agentic AI Engineer to design and build the LLM-powered agents behind their discovery platform. Your work will directly shape which therapeutic ideas move forward for patients who currently have no options.

What you’ll be doing

• Designing and building LLM agents that help scientists generate, prioritise and justify testable therapeutic hypotheses
• Working with drug discovery scientists to understand their real problems, and judging where an agentic approach adds value and where it doesn’t
• Connecting agents to a biomedical knowledge graph, in-house analytical methods and the scientific literature so they reason over the right evidence
• Extending and maintaining the existing suite of GenAI tools as the team’s needs and the wider ecosystem evolve
• Helping set up evaluation, testing and quality standards for agentic systems
• Writing clean, well-tested, well-documented Python that others can build on

The setup

• An established Python-based agentic framework is already in place, with workflows in production
• You’ll report to senior technology leadership and work alongside the technical lead for agentic infrastructure
• You’ll sit in a cross-functional team of ML engineers, software engineers, bioinformaticians, drug discovery scientists and clinicians
• There’s clear scope to take ownership of workflows and grow your influence over time
• By month 3: You know the framework and the scientific problems it serves, you’ve shipped your first contributions, and you’re turning scientist feedback into improvements.
• By month 6: You’ve taken at least one agentic workflow from concept to a production tool used in the discovery pipeline, and you’re owning workflows end to end.

What you’ll bring

• 2+ years’ experience as a software engineer or ML engineer
• A track record of building and shipping LLM-based agentic systems in production, including tool use, multi-step orchestration and reliable behaviour
• Strong Python and solid engineering fundamentals
• Hands‑on knowledge of the modern LLM stack: model APIs, prompting, tool calling, RAG, MCP, agent frameworks and evals
• Enjoyment of working closely with non-technical colleagues to understand what they need

Nice to have

• Experience in drug discovery, biology or another life sciences field
• Experience with knowledge graphs or reasoning over structured and mixed data sources
• Experience building evaluation harnesses or test strategies for LLM systems
• A habit of getting up to speed quickly in unfamiliar domains
• A genuine interest in techbio and patient impact

#J-18808-Ljbffr