Postdoctoral Appointee - AI for Biomedical Discovery

About the role

The Argonne team is seeking two highly motivated postdoctoral researchers to help shape the next generation of secure, scalable, and continuously learning AI systems for biomedical discovery. This position will contribute to the Forge project, which is focused on developing advanced multimodal AI capabilities that can learn across distributed data environments without requiring sensitive data to be centralized.

Responsibilities

• Conduct research and development in federated learning, privacy-preserving machine learning, multimodal AI, and foundation model adaptation for biomedical and related scientific applications.
• Develop new methods for multimodal federated learning that can integrate information across distributed datasets, including imaging, omics, clinical, text, sensor, and other structured or unstructured data modalities.
• Design and implement continuous learning approaches that allow models to improve over time as new data, validation results, or experimental feedback become available.
• Explore agentic AI approaches for federated learning, including AI agents that can assist with task orchestration, experiment planning, model evaluation, workflow automation, and decision support across distributed environments.
• Build and extend software capabilities in federated learning frameworks, with emphasis on scalable, reproducible, secure, and extensible research software.
• Evaluate model performance, robustness, generalizability, fairness, privacy, and data readiness across heterogeneous sites and datasets.
• Contribute to the design of secure AI workflows that may involve trusted research environments, secure enclaves, privacy-preserving computation, differential privacy, secure aggregation, or related techniques.
• Collaborate with interdisciplinary teams, including AI researchers, biomedical scientists, software engineers, security experts, and high-performance computing specialists.
• Prepare research results for publication in peer-reviewed conferences and journals, and communicate findings through presentations, technical reports, project meetings, and software documentation.
• Support project milestones, demonstrations, and deliverables by developing working prototypes, experimental benchmarks, and reusable software components.

Requirements

• Ph.D. completed within the last 0–5 years in computer science, data science, biomedical informatics, computational biology, bioengineering, applied mathematics, electrical engineering, or a related field.
• Strong programming skills in Python and experience developing research or production-quality machine learning software.
• Experience with machine learning or deep learning frameworks such as PyTorch, TensorFlow, JAX, or similar tools.
• Knowledge of federated learning, distributed machine learning, privacy-preserving AI, foundation models, multimodal learning, continual learning, or related areas.
• Ability to design and conduct computational experiments, analyze model performance, and communicate results clearly.
• Experience working with large-scale or complex datasets, including structured, unstructured, multimodal, biomedical, scientific, or high-dimensional data.
• Ability to work independently while contributing effectively to a multidisciplinary research team.
• Strong written and oral communication skills, including the ability to prepare manuscripts, technical reports, presentations, and documentation.
• Ability to model Argonne's core values of impact, safety, respect, integrity, and teamwork.

Preferred Skills and Qualifications

• Experience developing or extending federated learning frameworks such as APPFL, Flower, FedML, NVIDIA FLARE, or similar systems.
• Experience with multimodal biomedical data, including combinations of clinical records, medical imaging, pathology, genomics, transcriptomics, proteomics, wearable/sensor data, or scientific text.
• Familiarity with foundation models, large language models, vision-language models, biomedical AI models, or model fine-tuning methods such as LoRA, adapters, instruction tuning, or retrieval-augmented generation.
• Experience with continual learning, active learning, reinforcement learning, closed-loop learning, or human-in-the-loop AI workflows.
• Experience with agentic AI frameworks, tool-using LLMs, workflow orchestration, AI planning systems, or multi-agent systems.
• Familiarity with privacy and security techniques such as differential privacy, secure aggregation, secure multiparty computation, homomorphic encryption, trusted execution environments, or secure enclaves.
• Experience with distributed computing, cloud computing, containers, Kubernetes, Docker, Apptainer/Singularity, or high-performance computing environments.
• Experience with MLOps, reproducible workflows, experiment tracking, CI/CD, software testing, benchmarking, or open-source software development.
• Familiarity with biomedical AI validation, data readiness assessment, model evaluation, regulatory-grade evidence generation, or independent verification and validation workflows.
• Demonstrated ability to publish research, contribute to collaborative software projects, or present technical work to interdisciplinary audiences.