The research direction
Behavior in context.
Accountability across boundaries.
Models, agents, tools, and services influence one another. A decision can move through several systems, leave partial evidence, and create effects beyond the place it began.
My independent protocol development and AI security research investigates anomaly detection, provenance graphs, and blockchain accountability. The work explores how evidence can help trace activity across connected systems.
These are research questions under investigation through literature, prototypes, and experiments.
01 Provenance & evidence
Investigating what can be reconstructed from the lineage of data, decisions, and execution across interacting systems.
02 Blockchain accountability
Exploring tamper-evident records of origin and activity, and their relationship to the evidence collected by a system.
03 Anomalies & system behavior
Connecting statistical signals, learned representations, and structural context to the study of unusual behavior.