About the Role
We are seeking an expert AI Scientist to join the Data Research Department at the Samueli Cancer Research Institute. You will play a pivotal role in our mission to transform clinical care through advanced data science. In this role, you will lead the development of robust, scalable workflows for extracting complex medical insights from unstructured clinical text, managing the lifecycle of Large Language Models (LLMs) in a highly secure, research-focused environment. You will be responsible for bridging the gap between cutting-edge NLP research and clinical application, ensuring that our models meet the highest standards of accuracy and reliability in a medical context.
What You'll Do
- Extract medical data: design and develop pipelines that extract structured medical information from unstructured free-text clinical records.
- Develop & apply LLMs: apply, fine-tune, and evaluate a range of large language models for clinical information extraction, selecting the most suitable approach for each task.
- Validate model outputs: define and implement methods for validating and quality-assuring model results, including evaluation frameworks, error analysis, annotation strategies, and metrics for accuracy and clinical relevance.
- Embeddings & vector search: build embeddings and vector-database solutions over clinical text to power semantic search, retrieval-augmented generation (RAG), and similarity-based matching.
- Cross-disciplinary collaboration: translate complex clinical/oncological requirements into technical AI solutions, working closely with researchers and medical staff.
What We're Looking For (Must-Haves)
- Degree in Computer Science, Data Science, Bioinformatics, or a related quantitative field.
- 4+ years of proven, hands-on experience working with multiple LLMs (prompting, fine-tuning, evaluation, and integration into pipelines).
- Proven hands-on experience in machine learning model development and deep expertise with LLMs and NLP libraries (e.g., Hugging Face, PyTorch, LangChain).
- Strong background in statistics and evaluation metrics (AUC, F1-score, precision/recall, bias assessment).
- Strong end-to-end data workflow capabilities, covering data extraction (SQL or similar query languages), data cleaning, statistical analysis, and communication of insights.
- Hands-on experience with embeddings and vector databases, including embedding generation, indexing, and similarity search.
- Excellent Python skills and familiarity with relevant ML/NLP frameworks.
- Comfort working with noisy, real-world clinical data.
- Strong English communication skills and an ownership mindset: able to drive projects forward, prioritize, and follow through.
Nice-to-Have
- Experience working with medical text and healthcare data.
- Familiarity with clinical terminology and the challenges of sensitive, regulated medical data.
Why join Samueli?
Unmatched data access
Work with one of the largest healthcare databases globally.
Direct clinical impact
Influence care for ~1 in 7 cancer patients in Israel.
Startup energy, medical depth
Fast iteration with world-class infrastructure.