Kempner AI Fellows Program

The Kempner AI Fellows Program is for early-career researchers ready to contribute to ambitious AI/ML research. Fellows work with Kempner investigators on active projects spanning foundation models, AI systems, NeuroAI, computational neurobiology, and cellular and protein computational biology. 

The program is designed for early-career researchers with strong technical preparation in modern machine learning. Fellows deepen their research experience through hands-on work in collaborative scientific projects.

Program summary

The Kempner Institute for the Study of Natural and Artificial Intelligence at Harvard University seeks early-career researchers to join the Kempner AI Fellows Program.

AI Fellows work closely with Kempner faculty, researchers, students, and one another on research projects at the intersection of machine learning, neuroscience, and scientific applications of AI. Depending on the position, fellows may contribute to foundation model development and evaluation, agentic workflows and tool-using models, neural data modeling, brain foundation models, protein structure and docking, cell-state prediction, or multimodal biological modeling.

Fellows are expected to make substantive intellectual and technical contributions to active research projects. This may include developing research ideas, designing experiments, analyzing results, building or adapting modern AI/ML models, working with large-scale datasets, developing benchmarks or evaluation methods, and contributing to research outputs such as papers, code, models, datasets, or technical writeups. The role is especially well-suited to candidates who are eager to deepen their research skills while working in a highly collaborative, interdisciplinary environment.

Current positions

There are three AI Fellow positions, each with a distinct research focus and background expectations:

Fellow in AI/ML for Scientific Applications, Foundation Models, and AI Systems

• This position is for candidates with strong foundations in modern machine learning and the ambition to advance foundation models, agentic workflows/tool-using models, and AI systems for high-impact scientific applications (including the life sciences).
• Ideal candidates will have experience with: foundation model training, evaluation, adaptation, or post-training; agentic workflows; tool-using models; retrieval systems; large-scale datasets; distributed training; high-performance computing; or systems-level approaches to AI/ML. 

Fellow in AI/ML for NeuroAI and Computational Neurobiology

• This position centers on computational neurobiology and the use of modern AI/ML methods to model brain circuits and neural activity. Fellows may contribute to the development of brain foundation models that predict patterns of neural activity from large-scale, multi-regional recordings.
• Ideal candidates will have experience with: computational neurobiology or neural data analysis; time-series modeling; sequential models (transformers, autoencoders, dynamical systems); modeling neural activity from large-scale recordings.

Fellow in AI/ML for Cellular and Protein Computational Biology

• This position focuses on applying modern AI/ML methods to protein and cellular biology, including topics such as protein structure prediction, docking, cell-state prediction from large-scale perturbation datasets, and multimodal modeling of protein function and cellular state. 
• Ideal candidates will have experience with: computational biology or biological data analysis; protein structure modeling and docking; cell-state modeling from perturbation datasets (including Perturb-seq or related data); multimodal modeling for protein/cell tasks; familiarity with AlphaFold/RFDiffusion/CellCap or related systems.

Fellowship experience

AI Fellows contribute to ambitious, mentored research projects at the frontier of AI/ML, neuroscience, and scientific applications of AI. Fellows work closely with Kempner investigators and research teams to develop research questions, design experiments, analyze results, and contribute to new ideas, methods, models, and scientific understanding.

Depending on the project, fellows may:

• Help shape research directions, hypotheses, and experimental approaches.
• Apply machine learning to challenging questions in neuroscience, computational biology, foundation models, or AI systems.
• Develop, train, evaluate, or adapt modern AI/ML models as part of broader research efforts.
• Analyze complex datasets, interpret model behavior, and assess results against relevant scientific or technical questions.
• Communicate their work through research presentations, scholarly writing, technical documentation, and collaboration.
• Contribute to reproducible and open research outputs, including papers, code, models, datasets, benchmarks, or other shared resources.