FIELD NOTES / RESEARCH / CONVERSATIONS
Ideas for the learning enterprise.
Essays, field observations, and conversations about building AI systems that carry real work, preserve human judgment, prove their economics, and improve through the work.
FIELDRUNTIME / INSIGHTS
THE LATEST
From deployment to operating advantage.
The architecture around the model is the advantage.
OWNERSHIP / OPERATING LEVERAGE
Own the Leverage
The AI advantage is already inside your service business. Put AI around your experts and let operating knowledge compound inside your company.
LEARNING THROUGHPUT / ENTERPRISE CONTEXT
The Enterprise Bottleneck Is Learning Throughput
Why the companies that convert real work into verified context, evaluations, and reusable skills fastest will own the AI advantage.
EXECUTABLE JUDGMENT / ORGANIZATIONAL LEARNING
When Judgment Becomes Executable
How organizations turn accumulated human judgment into governed, reusable capabilities that improve through verified work.
LOOP ENGINEERING / AGENT OPERATIONS
Knowing When to Stop: The Art of Making a Loop Converge
Why production agents need verifiers, local repair, stop contracts, budgets, and economics that distinguish useful iteration from expensive motion.
AGENT SECURITY / TECHNICAL CONTROL
Agent Security Must Move From Process to Technical Controls
Why agent authority must be enforced through identity, permissions, isolation, policy gates, evidence, and tested recovery.
OPERATIONAL ASSURANCE
Why Enterprise Agents Need Operational CI
Business workflows need the agent-era equivalent of tests, sandboxes, versioning, staged release, monitoring, and rollback.
CONTINUAL LEARNING
The Harness Is Where the Company Learns
How production traces, corrections, failures, and outcomes become controlled improvements without changing model weights.
LEARNING ENTERPRISE ARCHITECTURE
A Company Brain Is Not Enough
Memory becomes an operating capability only when it connects to coordination, controlled action, verified outcomes, and learning.
THE DEPLOYMENT GAP
The Model Is Not the Deployment
A capable model becomes production infrastructure only through context, controlled action, verification, ownership, and learning.
COORDINATION RUNTIME / PRODUCT ARCHITECTURE
The Company That Knows What to Do Next
An open architecture for turning fragmented company activity into governed execution—and a system that learns from the work.
THE LEARNING ENTERPRISE
The Company Is Not a Brain. It Is a School.
Why enterprise AI will be built as specialized intelligences apprenticed to the people whose judgment makes them useful.
DECAGON / LEARNING SYSTEMS
The Model Is Not the Product
Ten operating rules for turning a capable agent into a controlled, measurable system that compounds in production.
ARCHITECTING THE AGENTIC ENTERPRISE
The Operating Threshold
Ten production rules for turning capable agents into trusted, governed systems that the enterprise can authorize to act.
GRAPH ENGINEERING
10 Rules for Enterprise AI That Works
A CIO's guide to building the workflow, state, checks, permissions, and learning system around the model.
WORKFLOW OPERATING SYSTEMS
Build the Operating System Around the Work
How learning workflow systems remove organizational weight, protect human judgment, improve execution, and turn everyday work into a compounding enterprise advantage.
COMING NEXT