Your operating model, readable by every AI your people use.
“Can every AI in the company answer from how we actually run — and stay inside the rules?”
Work with Dotwork from Claude, ChatGPT, Copilot, Cursor, Slack or Microsoft Teams. Every assistant reads the same governed model — strategy, money, portfolio, architecture, controls and the decisions between them — signed in as the person asking, with their permissions. The interface is whatever your people already use. The answers come from how the company actually runs.
Which strategic bets lost funding this quarter, and what did it cut?
From your Dotwork model
Two bets were cut back in the October reforecast. Expand mid-market lost $1.8M, pausing two onboarding initiatives. Self-serve analytics lost $0.6M, moving its launch a quarter.
Illustrative view · sample data
The pressure
Every AI tool in the company is guessing how it runs.
Copilots read documents, tickets and chat. None of them know the strategy in force, the budget behind it, the portfolio carrying it, the controls it has to meet or the decisions already made. So each AI answers the same question differently — and confidently.
What it looks like today
- The same question, a different answer in each AI tool
- Drafts built on last quarter’s deck instead of this month’s numbers
- Agents that can read the ticket but not the budget behind it
- Every new AI tool another integration to secure
What changes
What changes.
Context pasted into prompts by hand
The operating model, read live by the AI your people already use
Every AI tool with its own copy of the truth
One governed model behind every assistant and agent
Logging in to find the answer
Answers in Claude, ChatGPT, Copilot, Slack or Teams — linked back to the record
AI as one more integration to secure
AI that signs in as the person and sees only what they can
What AI can read
Not your documents. Your operating model.
AI without the operating model is confident guessing. Through Dotwork, every assistant reads the same connected model — and the relationships between its parts, so it can follow a question from strategy to spend.
Strategy
Bets, beliefs, objectives and key results — and what each one is meant to move.
Money
Budgets, forecasts, actuals and allocations, tied to the work they fund.
Portfolio and work
Initiatives, epics, dependencies and status updates across every team.
Architecture
Applications, capabilities and what depends on them, with end-of-life dates.
Controls and risk
Standards, controls, waivers and the systems they cover.
Decisions and history
Discussions, decisions and how every field changed over time.
…and how each part connects to the others, with the history of every change.
Where people work
Three ways in. One model behind all of them.
People ask from the AI they already use. Dot reaches them in the chat they already live in. Both work from the same model.
Ask and act from your AI
Any MCP client
Connect Claude, ChatGPT, Cursor, VS Code with GitHub Copilot, Claude Code or Codex to Dotwork’s remote MCP server. Sign in with your Dotwork account, pick your site, and your assistant can search the model, follow relationships and read the history behind any record.
- Remote server — nothing to install
- Search, relationships, trends and history
- Every answer links back to the record in Dotwork
Dot where your teams talk
Slack
Dot posts to people and channels in Slack — the update that needs an owner, the survey that needs an answer, the item waiting on an approval in Dotwork. Reply in the thread and Dot picks up where it left off.
- Messages to people and channels
- Replies in the thread continue the conversation
- Links straight back to the record in Dotwork
Dot inside your tenant
Microsoft Teams
The same conversations in Microsoft Teams, delivered through a bot registered in your own Microsoft tenant, with cards that link back to the record.
- Runs on a bot in your own tenant
- Cards to people, with replies that continue the run
- The same model behind every message
Ask it anything about how you run
The question gets asked where the person already is.
CFO · in ChatGPT
“Where are we over budget this quarter, and which strategic bets does it touch?”
Portfolio lead · in Claude
“Which initiatives have stalled, and what depends on them?”
Enterprise architect · in VS Code with Copilot
“Which applications reach end of life next year, and which capabilities run on them?”
Risk lead · in Claude
“Which control waivers expire soon, and on which systems?”
Team lead · in Slack
“Dot asks for this week’s status on two initiatives. A reply in the thread updates them.”
Product manager · in Cursor
“What did we commit to this quarter, and what changed since the plan?”
From reading to acting
When you’re ready, AI can update the model too.
Reading comes first. When a site’s admins switch on write access, assistants can keep the model current from wherever people work — and they show a draft before they change anything.
With write access on, an assistant can
- Create records and update fields
- Add status updates and comments
- Link work to the strategy, budget or control it belongs to
- Send a message to the right person in Slack or Teams
Every change is made as the signed-in person, within their permissions.
Governed by default
Governed like a person, because it acts as one.
Opening the model to AI shouldn’t open a new hole. Every assistant works through the same sign-in, permissions and site rules as the person using it.
Signs in as the person
Your Dotwork sign-in, with an explicit consent step — no shared service account.
Sees only what they can
Every call is checked against the person’s permissions — in planning, down to the cost centers and sensitive columns they may see.
One site at a time
Each connection is bound to a single Dotwork site, so nothing crosses between them.
Writes off until you turn them on
Write tools stay off for a site until its admins enable them, and admins can switch off any tool for a site.
Drafts before changes
Assistants are told to show a draft and ask before any change, and destructive tools are flagged so clients confirm first.
Always back to the source
Answers carry links to the records they came from, so anyone can check the AI’s work.
How it compares
Other tools hand AI their records. Dotwork hands it the operating model.
Work tools expose their tickets and documents to AI. Enterprise search indexes everything it can reach. Dotwork exposes how the company actually runs — strategy, money, portfolio, architecture, controls and the decisions between them — as one governed model.
| Capability | Work tools with MCP | Enterprise AI search | Dotwork headless operations |
|---|---|---|---|
| Use it from the AI tools people already have | |||
| Signs in as the person, with their permissions | |||
| Strategy, money, portfolio, architecture and controls as one model | Partly | ||
| Relationships and history, not just documents | Partly | Partly | |
| Trace a question from strategy to spend | |||
| Answers linked back to the source record | |||
| Writes off by default, switched on per site | Partly | ||
| Dot in Slack and Teams, on the same model as every AI client | Partly | Partly |
The model underneath
Every module makes the next one smarter.
Steer, Fund & Optimize and Govern & Deliver all read and write the same Enterprise Operating Model. The first module you license builds the model. Every module after it adds context instead of another silo — and Dotwork AI reads all of it, stopping where the decision is yours.
See the Enterprise Operating ModelBe Decisive.Build Fast.Connect Everything.
Start ConnectingSolutions
Platform
Resources
© 2025-2026 Dotwork, Inc. All rights reserved.