The Enterprise Operating Model is the living model underneath every Dotwork solution — your strategy, what the business does, the work underway, the rules it follows and what it all costs, connected in one place. Dotwork AI reads it continuously, flags what needs a decision and leaves the judgment to you.
The enterprise becomes self-steering. Never autonomous.
Three layers
Strategy, the business and the work, in one connected model
Six steps
From a strategy to a real dollar spent — and back
One set of numbers
Every dollar matches what finance actually booked
Where this comes from
It arrived one sensible decision at a time. Finance got its system, delivery got another, cloud got a third — each right about its own area and blind to the others.
Then came the tools that tried to see across them: a project portfolio tool, a cost tool, a cloud dashboard, an architecture inventory, an AI register. Each added a view. Each brought its own way of describing the business and its own integration to maintain. Connecting them stopped being a few weeks a quarter and became a permanent job.
Then AI arrived, asking for the one thing none of that produced. An agent is only as useful as the context it can reach. Rolled out department by department into the same silos, it inherits the same disconnect.
Connected, the same enterprise looks like this: every area of the business on one model, joined by the data it shares.
Reference models, then your model
You do not start from a blank page or a year-long design project. Dotwork comes with reference models — the structure of strategy, capabilities, work, rules and cost that most enterprises share — plus industry editions for banking, insurance, healthcare and construction. They are a starting point, not a template you have to fit.
The differences that make your enterprise yours — your structure, your terms, your approval paths, your cost rules — are modeled on top, as far as you want to take them. Go live on the reference and refine as you learn.
01
Start from the reference
A working model on day one, not a design project.
02
Fill it with your data
Your systems populate the reference with your real numbers.
03
Shape the last mile
Rename, add and adjust until it matches how you operate.
How your model is built
Your last mile
Your structure, terms, approval paths and cost rules
Industry edition
Banking, insurance, healthcare or construction
Reference model
Strategy, capabilities, work, rules and cost most enterprises share
Most of the model is ready on day one. The part that is uniquely yours is where your effort goes.
From there to here
A system for each department, right about its own area and blind to the others
The same systems, kept as the official source, connected into one model
Connecting and reconciling data as a permanent job
The connections kept up to date automatically
A company-wide picture rebuilt every quarter, out of date on arrival
One model that updates as your systems do — you look it up instead of rebuilding it
An operating model designed from scratch, months before anyone can use it
A reference model on day one, with the last mile shaped to how you actually run
A multi-year transformation plan, obsolete before it lands
Change made one capability at a time, measured by how fast you respond
Every department with its own numbers, and an argument about whose are right
One set of numbers, with every department seeing its own view of it
AI rolled out department by department, without the context it needs
AI reading the same model, following the same rules, with its cost counted in the same total
Three decision layers, one set of numbers
Direction sets where the enterprise competes and what it funds. Context turns that into the capabilities the business needs and the way value reaches customers. Execution delivers through governance, technology and people. Underneath all three sits one set of numbers that matches finance’s books — so what comes back up is a cost the CFO recognizes, not a slide.
Direction
The big choices: strategy and business model, what counts as success, which opportunities to pursue and where to invest. Every choice becomes an investment you can trace.
What comes back · What is feasible, what it costs and how it is performing flows back to inform the next decision.
Ready-made views on this layer
Context
The things the business must be good at, and the end-to-end flows that deliver value to customers. Without this layer, strategy jumps straight to projects and nobody can say what actually got better.
What comes back · Results, efficiency and AI effectiveness flow back to sharpen what the business focuses on.
Ready-made views on this layer
Execution
Oversight, technology and the people doing the work. This is where IT runs — and where most of the money is spent.
What comes back · Delivery and compliance data show the real consequences of each decision.
Ready-made views on this layer
One set of numbers
Every dollar comes in once, from what finance actually booked, and is used at every layer — never counted twice. Costs nobody could assign stay visible. Plans and contract commitments sit alongside actual spend, never mixed into it.
What comes back · What things actually cost flows back to the choices that called for them.
Ready-made views on this layer
Six steps, read in both directions
Left to right, you see why money is being spent: every funded bet tied to the goal it serves and the capability it improves. Right to left, you see what it cost: the real spend, traced back to the decision that asked for it. The same data answers both — so the strategy review and the budget review stop arguing about whose number is right.
Strategy
Where to compete
Goal
What must change
Investment
What is funded
Project
What is being done
Capability
What it improves
Cost
What it costs, matching finance’s books
A different question: did it work?
Every recorded decision keeps its expected result beside the actual one, on the measures it promised to move. Knowing the cost helps — but a decision is judged on whether it delivered what it promised.
Self-steering, not autonomous
The loop runs on the platform, not in a meeting. Dotwork AI watches the model and brings forward what needs attention. People judge every flag. And reviewing a flag is separate from approving a plan: nothing the AI notices becomes a plan on its own.
01 · Dotwork AI
Read
AI agents review the model on a schedule, within the limits you set.
02 · Dotwork AI
Surface
Anything worth attention is flagged, with how serious it is and the evidence behind it.
03 · Human
Judge
A person accepts, dismisses or corrects each flag. Anything the AI is unsure about is marked for review.
04 · Human
Approve
Only an approved plan goes live. If the numbers changed underneath it, you are warned — nothing is silently overwritten.
And at any time, anyone can ask the model a question in plain language — and read the steps behind the answer.
What it delivers
Insights
What becomes visible
Decisions
What can be decided — and revisited
What it sits on, and what it replaces
The practices stay: IT financial management, cloud cost management, portfolio management, enterprise architecture, AI governance. Your core systems stay the official source. What changes is the foundation those practices run on — and the manual reconciliation work that has been holding them together.
It sits on
Your core systems
The people who work in them keep working in them. The model reads them on their schedule.
It replaces first
The manual reconciliation
It lets thin out
Single-purpose tools
A tool retires only once Dotwork covers what it does — on your schedule.
From complaint to answer
The problems every operating-model review names — duplicate projects, unapproved software and AI, unclear accountability — stop being complaints and become questions the model can answer, with the numbers to back them up.
Duplicate projects
Which projects are improving the same capability, and where do they overlap?
Unapproved software (shadow IT)
Which applications have never been checked against our standards, and what do they cost?
Unapproved AI (shadow AI)
Which AI tools in production have no record, and what did they cost?
Technical debt
Which debt is funded, which has an exception, and when do the exceptions expire?
Nobody accountable for results
Did the decision actually move the measure it promised to?
Competing priorities
What mix are we actually funding, compared with the mix we said we wanted?
Approval bottlenecks
Which decisions are being escalated higher than their risk calls for?
Vendor lock-in
Which capabilities depend on a single supplier, and when does the contract renew?
What breaks when this changes
Executive requests make sense where they are made and get expensive where they land. With a connected model, the capabilities, applications, teams, controls, contracts and past decisions a change affects are on the table before it is approved. The trade-off stays human — and it should.
How you would know it is working
None of them is a delivery measure. All of them come straight from the model, not from numbers somebody typed in.
More spend tied to strategy
The share of total spend that links to a strategic goal, going up.
Less unassigned cost
Costs nobody could assign, going down — with the missing detail named for each one.
Decisions that delivered
The share of recorded decisions whose promised result actually happened.
Time to respond
How long it takes from something being flagged to someone acting on it — timestamped on both ends.
Three solutions, one model
Start with the solution where the pressure is highest. Each one you add enriches the same model instead of creating another silo.
Choose
What should we do — and why?
Strategy continuously calibrated to reality.
Explore SteerFund
Where is our money going — and is it going to the right places?
Maximize the value of every dollar spent.
Explore Fund & OptimizeDeliver
Is the work getting done — the right way, and for the right reasons?
From plan to proof, in one continuous learning loop.
Explore Govern & DeliverBefore you buy
How to get a number instead of a pitch.
1
Good answer · Goes from the strategy through the work to a figure that matches finance’s books.
Bad answer · Ends at a total somebody typed in.
2
Good answer · Lists what is unassigned and the missing detail behind each item.
Bad answer · Claims everything is assigned.
3
Good answer · Opens a record with the opportunity, the project and the result it promised.
Bad answer · Opens a slide deck.
Walk the loop
A 45-minute working session on the Enterprise Operating Model. Pick a strategy, follow it to the cost it actually consumes, and follow the cost back.
Request a working session© 2025-2026 Dotwork, Inc. All rights reserved.