Published and ongoing research, with code and papers where available. The full list of papers is under Publications.
GraphBit · EMNLP 2026 Industry Track
GraphBit models agents, tools, and memory as nodes in an executable graph, enabling branching, parallel, and cyclic multi-agent workflows. I lead the project and designed its orchestration model and execution engine.
Ongoing Research · MTSU
LLM agents that plan drone missions and learn better navigation strategies from experience, combining a Markov decision process formulation with model-predictive control. I lead the reinforcement learning and MPC modeling.
Ongoing Research · MTSU
Deep architectures for hyperspectral data cubes with scarce labels, from multidimensional CNNs, residual attention, and dynamic kernels to energy-based generative models. I lead the project, designing and implementing the 3D modeling.
Ongoing Research · MTSU
Graph learning models for node classification and link prediction that quantify uncertainty, flagging predictions on unfamiliar nodes or edges that should not be trusted. I lead the graph modeling work.