Research

Navrysa Labs

Navrysa Labs explores emerging problems at the intersection of cybersecurity, artificial intelligence, autonomous systems and the physical world. It is the research and experimental development arm of the company.

Research areas

Four directions, one question.

Everything here follows from the same question: how does a machine judge a situation well enough to be trusted with a consequence?

Physical Intelligence

Context-aware perception

Context-aware perception, temporal reasoning, risk assessment and constrained response for systems that observe the physical world.

Cybersecurity Research

Behavioural defence

Runtime integrity, provenance-based security, timing anomaly detection, cyber-physical and IoT security, and AI-enabled cyber defence.

Trustworthy AI

Assurance for intelligent systems

Secure AI systems, LLM and agent security, AI trust, runtime assurance and AI system integrity.

Autonomous Systems

Resilient autonomy

Resilient autonomy, adaptive perception, resource-aware intelligence, security for autonomous platforms and cyber-physical resilience.

A dark engineering workbench lit by a single warm task lamp: a camera module on a machined mount, a bare circuit board trailing a ribbon cable, a partly disassembled metal enclosure, and hand tools laid out on the surface.
Prototype benchConcept render

Collaborate

Research partners welcome.

Navrysa is interested in university, industry and government collaborations that align with this research agenda. Reach the research team directly at research@navrysa.com.