You will be at the forefront of developing a sophisticated scientific reasoning system that leverages Large Language Models (LLMs) to provide structured, biologically-grounded explanations. Collaborating closely with immunologists, machine learning researchers, and technical leadership, you'll shape how we derive insights at a systems level, pushing the boundaries of AI in biology.
Location: Ramat Gan, Israel (Hybrid role)
What will you do?
Design, prototype, and build LLM-based agentic systems that reason over biological data, scientific literature, model outputs, and internal tools.
Develop agents capable of structured reasoning, hypothesis generation, explanation, planning, tool use, and iterative scientific analysis.
Build robust evaluation frameworks for agentic systems, including automated and human-in-the-loop evaluation pipelines.
Define and implement benchmarks, metrics, and test suites for measuring agent performance, including reasoning quality, biological grounding, factuality, robustness, reproducibility, and usefulness.
Work closely with AI researchers, computational biologists, immunologists, and product teams to translate scientific needs into measurable AI capabilities.
Create evaluation datasets and benchmark tasks that reflect real-world biological and therapeutic reasoning problems.
Analyze agent behavior, failure modes, hallucinations, tool-use errors, reasoning gaps, and grounding issues.
Contribute to the architecture of production-grade AI systems, including agent orchestration, retrieval, tool calling, memory, planning, and monitoring.
Stay up to date with the latest developments in LLMs, agentic AI, evaluation methodologies, and scientific AI systems.
Help turn research prototypes into reliable products used by internal teams and external partners.
MSc or PhD in Computer Science, Electrical Engineering, Computational Biology, Statistics, Mathematics, or a related quantitative field.
Strong background in machine learning, data science, statistics, or computational modeling.
Hands-on experience building with LLMs and agentic AI systems.
Proven ability to design evaluation methodologies for AI systems, especially LLM-based or agent-based systems.
Experience working with LLM APIs such as OpenAI, Anthropic, Google, or open-source LLMs.
Experience with agent frameworks or orchestration tools such as LangGraph, LangChain, or similar systems.
Experience defining benchmarks, metrics, validation sets, scoring methods, or automated evaluation pipelines.
Strong Python skills and ability to write clean, production-aware research code.
Ability to work with complex, noisy, high-dimensional data.
Strong communication skills and ability to collaborate with experts from different disciplines.
