Active internal framework
OI SUITe
Human-centered operating and augmentation layer for AI workers: workrooms, authority, memory, telemetry, evidence, review, and continuity.
Human authority // systems architecture
Build the operating structure around the intelligence.
Identity, authority, policy, bounded execution, telemetry, verification, and receipts — with the human remaining responsible for consequential decisions.
Mason Perry
Founder, NULLWORKS · Operational Intelligence Systems Architect
Human-centered operational architecture for AI workers and consequential systems
About
I build governed operating structures around AI workers so capability does not quietly become authority.
I am the founder of NULLWORKS and a pioneering Operational Intelligence Systems Architect (OISA). My work sits between operations, systems architecture, AI governance, human factors, evidence, and implementation.
The operating idea is simple: the model is not the operating system. Organizations need identity, authority, intent, policy, bounded execution, telemetry, verification, continuity, and records of what actually happened.
The accidental origin story
I started using modern generative AI on February 6, 2026. I did not begin with an AI-governance theory. I treated the system like a digital employee and kept solving the operating problems that appeared: memory, handoffs, authority, correction, evidence, failure recovery, and continuity.
Only later did it become obvious that NULLWORKS itself had become an experimental organization for the question we were asking: what happens when AI stops being software an organization uses and starts becoming labor that an organization must organize?
Current work
NW
NULLWORKS
Designing the factory around AI workers: work cells, authority boundaries, continuity, quality gates, telemetry, provenance, and human control.
OPS
U.S. Postal Service
Front-line exposure to exception handling, logistics, operational friction, field systems, and the gap between how a process looks on paper and how work actually moves.
Selected systems
Operational infrastructure
NEURAXIS
Vendor-neutral authenticated ingress into governed NULLWORKS context. The current public-facing ingress is deliberately bounded and read-only.
Active research
PORTUS LUNARIS
Lunar infrastructure R&D: power, mobility, surface preparation, specialized robotics, interfaces, recovery, and logistics-first port architecture.
Published exercise
Operational assurance work
Technical pressure-testing and evidence-preserving assurance exercises. Published artifacts are explicitly not represented as certification or independent assurance.
Evidence, not vibes
NULLWORKS distinguishes between what exists, what has been tested, what has been deployed, and what remains research. The goal of this page is to do the same for the person behind the systems.
Model-agnostic continuity
Testing whether identity, doctrine, evidence, authority, correction history, and mission context can survive migration across model vendors while the reasoning engine changes.
Machine-verifiable governance
Current experiment direction: can an unfamiliar model reconstruct authority, a blocked action, revocation, recovery, and final state from raw operational receipts without being given the narrative first?
Continuity Calculus
A developing framework for preserving operational intelligence across threads, workers, models, interruptions, and correction cycles.