Associate Professor, University of Oregon
Chief Scientist, Link Oregon
Making digital infrastructure observable, intelligent, and resilient.
Modern digital infrastructure spans physical networks, clouds, programmable systems, submarine cables, satellites, and emerging quantum-classical systems. These layers often hide the dependencies that determine how systems behave and fail.
My research combines measurement, systems, and AI to reveal hidden dependencies, explain behavior from incomplete evidence, and engineer infrastructure that can remain resilient under physical failures, attacks, changing operating conditions, and increasingly autonomous operation.
We develop measurement and inference techniques that combine network observations with external evidence to uncover physical infrastructure, topology, shared dependencies, and other structure that cannot be observed directly.
We develop systems that connect traffic, routing, telemetry, DNS, active measurements, logs, topology, and infrastructure data to explain what changed, why it changed, and what evidence supports that explanation.
We design systems around hidden dependencies, resource constraints, attacks, physical hazards, and changing operating environments across programmable packet and optical systems, critical infrastructure, cloud and satellite networks, and emerging quantum-classical networking.
Our paper on capability-aware agent discovery for network operating systems was accepted to the NeurIPS 2026 AgenticOS Workshop. Congrats Riya and River!
Our paper on optimizing wildfire sensing networks with multi-agent reinforcement learning was accepted to the NeurIPS RL4XS Workshop. Congrats River!
Our paper on LEO networking research was accepted to ACM HotNets 2026. Congrats Chris and Zander!
Our work on integrating AI into computing curriculum will appear in SIGCOMM workshop!
Excited to receive an NSF CICI TCR grant to work on securing campus CI and REN workflows. Thanks NSF!