One Degree executive training

The Decision Velocity Enterprise

AI is making insight, software, and coordination faster. Most industrial operating models were designed for a slower world. Learn where decisions wait, what the waiting costs, and how to redesign authority, software, and organizational structure for faster controlled action.

Two-hour briefing · Three-hour masterclass · Four-hour working session · Custom executive cohort
The business problem

Industrial companies rarely lack information. They suffer from the distance between a signal and a legitimate response.

A machine begins to degrade. A supplier commitment becomes uncertain. Quality starts to drift. The organization detects the condition — then the evidence sits across systems, experts disagree, authority is unclear, and intervention waits for the next meeting. The loss created in that interval is decision leakage.

Signal detected Context assembled Decision prepared Authority located Action coordinated Outcome stabilized
What leaks while it waits

Downtime · Inventory · Expedites · Rework · Yield loss · Schedule churn · Margin erosion · Customer concessions · Executive burden · Lost intervention windows

Estimate the leakage

What is the waiting costing you?

The accounting system has no line called decision latency, so the cost is distributed across operations, supply chain, maintenance, quality, finance, and management time. This puts a first bound on it.

$ millions

A condition is developing in the plant or the supply base. How often is it understood only after it has already cost you something?

How often do people rebuild the same picture by hand, pulling evidence from systems that disagree with each other?

How often does action wait on an approval from someone who did not see the condition?

How often does a single decision cross three or more functions before anything changes on the floor?

How often does an intervention need several groups to coordinate before the operating condition actually changes?

How often does the same problem come back because the fix was never stabilized?

Estimated annual leakage

Three streams, nine value pools, from Michael Carroll's Bill Comes Due framework. Detection is priced against revenue at a 1.5% rate over a three-week average delay; decision against the 2.5% of revenue that moves through time-sensitive commitments; execution against operating expense at a 6% rework rate. The band is not a decorative margin — its width is the width of the benchmark ranges behind each pool. Six questions cannot produce a number worth defending to a CFO, and no floor is set at zero, because no operating system leaks nothing. Treat this as an order of magnitude and an argument for measuring the real one.

Five questions

Where is your constraint?

A short recognition exercise. Answer honestly — the result points at the likely source of latency, not a score.

1. Does your organization detect important operating conditions early enough to still have a choice?

2. Do the same decisions repeatedly cross several functions or approval levels?

3. Are senior experts routinely pulled into recurring exceptions?

4. Is AI producing more recommendations than the organization can govern or execute?

5. Can you quantify the cost between initial signal and stabilized response?

Likely primary constraint

This is a recognition tool, not a validated assessment.

Why AI changes the issue

Insight is becoming cheaper. Consequence is not.

AI does not automatically make the enterprise faster. It moves the constraint. When insight becomes abundant, decision rights, operating structures, controls, and accountability determine whether the company can act.

AI reduces the time required to
  • Assemble context
  • Analyze conditions
  • Generate options
  • Create software
  • Coordinate work
That moves the constraint toward
  • Authority
  • Permission
  • Organizational boundaries
  • Risk acceptance
  • Accountability
  • Outcome control
A recommendation generated in seconds can still wait days for permission.
What the course examines

Five things that decide how fast a company can act.

01

Decision latency

Where time enters between changing conditions and effective intervention.

02

Decision leakage

How waiting becomes downtime, inventory, freight, rework, lost margin, risk, and consumed management capacity.

03

AI and repricing

How cheaper intelligence, software production, and knowledge work move value toward context, judgment, integration, assurance, and outcomes.

04

Software as operating intent

How adaptive systems move from records and workflows toward bounded intervention and continuous learning.

05

Organizational redesign

How decision rights, management layers, functional boundaries, human roles, agent roles, and escalation structures affect velocity.

What participants leave with

One consequential decision, seen differently.

Not a generic AI framework and not another list of leadership principles.

Course structure

Ten modules, one decision path.

The three-hour masterclass runs the full sequence. Shorter formats compress it; longer formats add company-specific work.

01The Cost of Waiting
What does decision latency cost when the accounting system has no line called “decision latency”?
  • Decision latency versus slow decision-making
  • Visible losses and hidden leakage
  • The intervention window and reversibility
  • Why a correct decision made too late is still a poor decision
Output — a leakage hypothesis for one recent event
02Where Decisions Stall
Where does time enter the path between evidence and action?
  • The eight-stage decision path, from event to learning
  • Conflicting definitions, functional handoffs, approval layers
  • Meetings used as permission systems
  • Decision geometry — the hops and boundaries between signal and action
Output — a mapped decision path with the delay points marked
03Leakage and Capacity Loss
Why do capable organizations lack the capacity to improve?
  • The capacity cascade — from weak control to firefighting as the operating model
  • Capability is not capacity
  • Which experts function as human middleware
  • Where heroism is hiding weak architecture
Output — an inventory of escalations that have become accepted management work
Supporting reading — Entropy Rising ↗ · Chaos to Command ↗
04AI Moves the Constraint
What changes when analysis, software creation, and coordination become less scarce?
  • What AI accelerates — retrieval, context, options, coding, monitoring, recall
  • What it does not resolve — ownership, authority, risk acceptance, override
  • The AI latency paradox: becoming faster at producing undecided work
  • Observe, infer, recommend, prepare, execute — under what evidence and constraints
Output — a first pass at the observe-to-execute boundary for one decision
05The Repricing of Software and Services
What happens to software and advisory services when AI lowers the cost of producing both?
  • What made traditional software and services economics expensive
  • Where premium value moves — context, encoded method, judgment, assurance
  • What does not change — brownfield reality, exceptions, continuity, trust
  • Reading vendor and consulting claims against that shift
Output — a repricing view of current software and services spend
06Software as Operating Intent
What does software become when it can adapt closer to the moment of decision?
  • From strategy → policy → process → requirements → software, to intent → context → reasoning → bounded action
  • Five stages: record, workflow, recommendation, bounded intervention, adaptive operating layer
  • Why loading documents into a model is not an operating context
  • Information → reasoning → permission → action → learning
Output — a maturity placement for the systems behind your decision
07Organizational Consequences
Which parts of the organization exist because information, expertise, and software were once hard to move?
  • From functional routing to decision ownership
  • From management as aggregation to management as judgment
  • From roles alone to human-machine decision capabilities
  • From span of control to span of decision
Output — the roles your decision currently depends on, and why
08Permission in Advance
How can an enterprise move faster without accepting uncontrolled risk?
  • The permission ladder — observe, recommend, prepare, act in a bounded envelope, never
  • What expands authority — consequence, reversibility, evidence quality, demonstrated performance
  • Operating envelopes, escalation criteria, and human override
  • Decision journals and outcome records
Output — a draft permission ladder for one decision family
09Industrial Decision Families
Where does this bite hardest in an industrial operation?
  • Maintenance and asset health · quality and process deviation
  • Supply-chain disruption · schedule and customer commitment
  • Capital and engineering change
  • Product and service support — including non-human consumers
Output — the decision family worth redesigning first
10Executive Working Session
What would this decision look like if it were designed for velocity?
  • Define the decision, its frequency, and the value exposed
  • Map the current path and identify the leakage
  • Map current permission, then design the future path
  • Define success — signal-to-action time, handoffs, escalations, stabilization
Output — a completed first-draft Decision Velocity Canvas
Delivery formats

Built for the room you have.

Executive briefing
2 hrs

For boards and leadership teams

Latency, leakage, how AI moves the constraint, and permission in advance. Leaves a shared leadership vocabulary and candidate decision families.

Executive masterclass · recommended
3 hrs

The flagship format

The full sequence, including the decision-family exercise. Leaves a preliminary decision map, leakage hypothesis, permission assessment, and redesign opportunity.

Company working session
4 hrs

One company or business unit

A company-specific industrial case, breakout decision mapping, human-AI responsibility mapping, and a 30-day action plan. Leaves a first-draft Decision Family Charter.

Custom executive cohort. Delivered over several sessions for a leadership team — latency and leakage, AI and the repricing of work, software as operating intent, permission and bounded autonomy, operating-model redesign, an applied decision-family lab, and an executive review. Participants apply the work between sessions and return with operating evidence.
Supporting reading

The course sits inside a body of thinking.

Start with three. Then follow the argument into the Dispatch.

Read first

AI Advantage Is Not Intelligence — It Is Decision Architecture

The core claim the course is built on.

Read first

The Latency Tax

What the enterprise pays while a known condition waits for a decision.

Read first

Agents Don’t Create Value. Authority Does

Why capability without permission produces recommendation, not result.

Latency & controllability

The Latency Trap

How elapsed time, not meeting speed, decides the outcome.

Latency & controllability

The Cost Curve Starts Too Late

Why the cost of an event begins accruing before anyone books it.

Latency & controllability

Control Without Controllability

The difference between watching a process and being able to change it.

Authority & permission

The Permission Staircase

Bounded authority as a ladder, from observe to act to never.

Authority & permission

The Architecture of Permission

The permission structure every company is already running, unexamined.

Authority & permission

Control Is the COO

Where outcome ownership actually sits in the operating model.

AI & adaptive software

The World Model Is Necessary — But Is It Enough?

Why representation alone does not produce action.

AI & adaptive software

Adaptive Operational Architecture

What software becomes when it expresses current intent under current conditions.

AI & adaptive software

Composable Control

Assembling control from parts, without surrendering accountability.

The remaining distance

Bring the decision your organization keeps waiting to make.

The signal may already exist. The evidence may already be available. The expertise may already be inside the company. AI may be able to assemble the answer in seconds. The remaining distance is between knowing and acting — and that distance is designed.