Current work.

Compliance snapshots go stale the moment they are produced.

CAM is an R&D effort toward making security posture independently verifiable rather than asserted by dashboards. It converts fragmented security evidence into a continuously verifiable model connecting evidence, controls, assets, and risk.

Security testing produces conclusions faster than it produces evidence.

AEGIS explores coordinated AI agents for cyber inspection and defensive validation. Agents gather evidence, test hypotheses, and validate controls while maintaining a traceable evidence chain for every conclusion, under human supervision.

Robots see the world but barely feel it.

EPC is a robotics research architecture inspired by biological nervous systems: distributed peripheral sensing, locally encoded contact events, proprioception, and central predictive control, unified by an embodiment-independent somatosensory event schema.