Postdoctoral Fellow; PREP

Position Summary

Johns Hopkins University, Whiting School of Engineering, Office of Research and Translation. This role is part of the National Institute of Standards and Technology (NIST) Professional Research Experience Program (PREP). The associate will support the NIST FRAME (Foundational Representation and Assimilation for Multimodal Experiments) program by developing and validating generative AI and physics‑grounded modeling approaches that reconcile multimodal measurements for bioformulations and soft nanocarrier platforms.

Bioformulation Digital Twin Developer

U.S. citizenship is preferred but not mandatory.

• Develop, train, and validate generative models for 3D structure and mesostructure of soft matter systems, focusing on bioformulations and nanocarrier platforms.
• Design model architectures and training pipelines, including VAE, latent‑variable models, diffusion and score‑based models, autoregressive models, normalizing flows, or related approaches.
• Create representations that bridge cartoon or parametric structure generators, material digital twin representations, and experimental signatures such as SAXS, SANS, RSoXS, Cryo‑EM, and light scattering.
• Incorporate uncertainty quantification, calibration, and validation workflows so that model outputs can be compared rigorously with experimental observables.
• Define metrics and benchmarks for physical plausibility, diversity, reproducibility, and fidelity to measured data.
• Collaborate with experimentalists and instrument teams to close the loop between formulation, structure, measurement, analysis, and model update.
• Present results at internal meetings and occasional meetings with external stakeholders, including collaborators in measurement science, materials modeling, and user‑facility instrumentation.
• Produce open and reproducible research outputs, including documented code, datasets, metadata, model cards or equivalent documentation, protocols, and publications.
• Ensure that results, protocols, software, datasets, metadata, and documentation are archived or otherwise transmitted to the larger organization.

• Ph.D. completed by the start date in machine learning, computer science, physics, chemistry, materials science, chemical engineering, or a related field.
• Strong Python programming skills and experience with a modern machine‑learning stack, including PyTorch and GPU or HPC workflows.
• Demonstrated ability to execute independent research, communicate results, and publish in peer‑reviewed venues.
• Experience in generative modeling for scientific data is strongly preferred.
• Experience with generative AI for scientific or physical systems, including 3D fields, images or volumes, point clouds, graphs, or related structured representations, is highly desired.
• Experience with soft matter, self‑assembly, colloids, surfactants, polymers, biomaterials, or bioformulations is highly desired.
• Experience with inverse problems or simulation‑to‑measurement workflows, including learned forward models, differentiable physics, or amortized inference, is highly desired.
• Experience with scientific data engineering, including dataset versioning, provenance, metadata, or reproducible research workflows, is highly desired.
• Strong oral and written communication skills and ability to work collaboratively with experimentalists, instrument scientists, and computational researchers.

$85,000 - $100,000

The Johns Hopkins University is committed to equal opportunity for its faculty, staff, and students. The University does not discriminate on the basis of sex, gender, marital status, pregnancy, race, color, ethnicity, national origin, age, disability, religion, sexual orientation, gender identity or expression, veteran status, or other legally protected characteristics. The University is committed to providing qualified individuals access to all academic and employment programs, benefits, and activities on the basis of demonstrated ability, performance, and merit without regard to personal factors or demogrpahic characteristics that are irrelevant to the program involved.