Design knowledge-graph-augmented transformers and retrieval-augmented generation (RAG) pipelines that enable semantic querying and reasoning over materials-science/physics corpora, Developing pipelines for semantic enrichment of unstructured data, including entity recognition, relation extraction, and automatic ontology alignment in physics and materials domains, Build and maintain ontologies, OWL/RDF knowledge graphs, SPARQL endpoints, and open benchmarking suites to guarantee FAIR, reusable research data, Mine and link structure-property relationships from DFT, MD, phase-field, TEM/SEM, and other multimodal datasets from simulation and experiment, Develop benchmarking protocols and toolkits to evaluate AI models on materials science tasks as well as integrate your semantic-AI services into high-throughput GPU/HPC workflows, contributing to data management, metadata structuring, and semantic annotation, Collaborate with experimentalists and theorists to validate extracted knowledge via in-situ spectroscopy, synchrotron work, and high-throughput synthesis—and present your results at leading AI and materials conferences