Research Scientist - Multimodal Large Language Models (LLMs)

The Chan Zuckerberg Biohub San Francisco (CZ Biohub SF) (https://www.czbiohub.org/sf/) is an independent nonprofit research institute that brings together three powerhouse universities - Stanford, UC Berkeley, and UC San Francisco - into a single collaborative technology and discovery engine. CZ Biohub SF itself supports some of the brightest, boldest engineers, data scientists, and biomedical researchers to investigate the fundamental mechanisms underlying disease and develop new technologies that will lead to actionable diagnostics and effective therapies. We are guided by our values of scholarly excellence; disruptive innovation; hands-on engineering/hacking/building; partnership and collaboration; open communication and respect; inclusiveness; and opportunity for all.

Our Vision

• We pursue large scientific challenges that cannot be pursued in conventional environments
• We enable individual investigators to pursue their riskiest and most innovative ideas
• The technologies developed at CZ Biohub San Francisco facilitate research by scientists and clinicians at our home institutions and beyond

Diversity of thought, ideas, and perspectives are at the heart of CZ Biohub Network and enable disruptive innovation and scholarly excellence. We are committed to cultivating an inclusive organization where all colleagues feel inspired and know their work makes an important contribution.

The Opportunity

The Chan Zuckerberg Biohub (CZ Biohub SF) is seeking a highly skilled and motivated Research Scientist to lead the development of state-of-the-art multimodal large language model (LLM) agents that will enable breakthrough research and discoveries in biology. We are interested in pursuing these new ideas for zebrafish, a powerful model organism, to understand mechanisms of infection and immunity, organ regeneration, and organismal development.  The ideal candidate will have established expertise in machine learning, generative AI, development of multimodal LLMs, and reinforcement learning for model tuning. The successful candidate will report directly to both Yasin Şenbabaoğlu (Director of Computational Biology) and Loïc A. Royer (Director of Imaging AI) at CZ Biohub, San Francisco.

You will

• Design, develop, and implement multimodal LLMs that integrate textual, multi-omic, and image data.
• Lead the research and development of novel algorithms to process and align scientific literature with biological datasets for downstream analysis.
• Collaborate closely with computational biologists and experimental scientists to understand domain-specific challenges and optimize model performance.
• Manage large-scale datasets (scientific texts, omics data, and imaging) and build efficient data pipelines for training and evaluation.
• Drive innovative research to publish in top-tier AI and computational biology journals and present at conferences.
• Mentor junior scientists and engineers, fostering a culture of collaboration and continuous learning.

You have

Required –

• PhD in Computer Science, Machine Learning, Computational Biology, Bioinformatics or a related field; or Masters with equivalent experience
• 3+ years of experience with Python and relevant deep learning libraries (e.g., PyTorch, TensorFlow)
• 3+ years of experience in designing and deploying large-scale language models or multimodal AI systems
• Expertise in natural language processing (NLP), deep learning, and model training techniques
• Proven track record of impactful publications and conference presentations in relevant areas
• Excellent problem-solving skills and ability to work in an interdisciplinary environment
• Strong professional judgment and problem-solving abilities that adapt to a variety of situations
• Strong interpersonal skills with excellent written and verbal communication skills

Nice to have -

• Experience in integrating and aligning heterogeneous data sources (text, omics, images) for AI-driven applications.
• Familiarity with scientific literature databases (e.g., PubMed, arXiv) and bioinformatics