WC
The Operating System for the AI Era

Governance. Context. Control.

A body of work on how leaders build, govern, and orchestrate intelligent systems — from context engineering to agentic architectures to the transformation operating model.

Explore the WorkAbout the Book
The AI Executive's Handbook: Harnessing the Ungoverned Machine by Misha Sulpovar
The Discipline

The Framework

Four interconnected domains that define how organizations navigate the AI era. Each is a body of research, not just a topic.

Wired Cognition OS

AI Strategy

Product, organization, and transformation in the AI era.

Agentic Surface

Multi-agent architectures, control planes, and orchestration.

Knowledge Architecture

Writing, selecting, compressing, and isolating context at scale.

AI Governance

Frameworks, compliance, ethics, and the EU AI Act.

Latest Research

Recent Work

Also Recent
A governance layer deciding which artifacts are approved, current, and verified before retrieval runs
Context EngineeringFramework

What 'Context Governance' Actually Means

The plain-language version of our arXiv paper: the layer beneath retrieval that decides whether an artifact was approved, current, and verifiable before the agent ever reads it, and two experiments showing why that layer is cheaper and more correct.

3 min readOperational
A knowledge graph forming a brain over a terminal — a governed context frame for a second brain
Context EngineeringField Note

Your Second Brain Needs a Better Context Frame

Notes pile up, the AI forgets, and the whole thing turns to slop. The missing piece isn't a better model or another app — it's a context frame. Here's the one I built, on video, plus a guide you can hand to your own AI.

3 min readApplied
A hash-chained, version-verified context layer sitting beneath retrieval and an AI agent
Context EngineeringFramework

Context Governance, Now With a Paper

ContextNest is now a peer-reviewable spec. Our new arXiv paper formalizes context governance — the layer beneath retrieval that decides what's approved, current, and verifiable — and backs it with two controlled experiments.

2 min readOperational
Clinical AI moving from retrieval-only tools toward governed, frontier-model reasoning
Context EngineeringAnalysis

Healthcare Is Quietly Moving Past RAG

Specialized medical AI tools like OpenEvidence were supposed to beat general models by retrieving over peer-reviewed literature. In 2026 benchmarks, general frontier models are catching and passing them. The lesson isn't that RAG is dead — it's that retrieval was never the moat.

3 min readApplied
Organizational knowledge as connective tissue between people and an AI system — the missing context layer
AI StrategyAnalysis

Organizational Context Is Still the Bottleneck

Everyone is shipping agents. Almost no one is fixing the context they run on. A year into the pilot bonanza, the organizational context gap is exactly where it was — and it's why teams still feel behind.

3 min readApplied
An athlete in racing kit standing on a track beside officials in coats and high-visibility vests — the operator and the stewards, side by side
AI StrategyFramework

Athletes and Stewards

Most companies are arguing about whether to hire AI people or upskill domain people. There are two roles. And the second one — what most still call the domain expert — has quietly become a steward of the context the AI is actually reading from.

8 min readApplied
View all articles
Stay Connected

Executive AI intelligence,
delivered.

Governance frameworks, agentic architecture patterns, and strategic implementation insights. For leaders building the AI era.