How it works
A real work journey does not end when the conversation does.
MMT Orbit shows what it understood, binds evidence and decisions correctly, executes only authorized work and waits for the real outcome on the same record. The five phases on the home page group the ten steps below; both are shown on the same example job.
- 1 · GoalAn understanding summary2 steps
- 2 · CoordinationA list of sources and files3 steps
- 3 · ProductionA computation output and the file used2 steps
- 4 · EvidenceA draft response letter and change list1 step
- 5 · ValueA work history you can continue2 steps
Correcting the interpretation, the method decision and approving the response stay with the researcher.
One example job
Five phases, ten steps, one example
Select a phase to jump to its detailed steps. Each step shows what happens in the example job, your decision and the record created.
A reviewer report arrived: revising the manuscript
Illustrative example · starts no real work · not a measured resultA researcher received a reviewer report on a manuscript. One reviewer asks whether the finding still holds when a control variable is added. There are six weeks for the revision; the researcher has the report (PDF), the manuscript draft (DOCX) and the data table (CSV).
- 1GoalUnderstand the purpose, available inputs and limits.Steps 1, 2
- 2CoordinationBring information, tools and decisions into the same context.Steps 3, 4, 5
- 3ProductionPrepare the work and outputs through permitted tool paths.Steps 6, 7
- 4EvidenceSee sources, calculations, gaps and points needing review.Step 8
- 5ValueEvaluate the outcome for use and, where appropriate, the next decision.Steps 9, 10
1Goal
Understand the purpose, available inputs and limits.
- 01
Describe the situation
Write what happened, the outcome you want, the constraints and what you already have, in plain language.
In the example: “A reviewer report arrived; there is a control-variable question and I have six weeks.” The report, draft and data file are added to the work.
- 02
See what MMT understood
Goal, current state, knowns, unknowns and uncertainty appear as a small proposal.
In the example: Goal: revision. Known: report, draft, data. Unknown: the journal's current rules and whether the control variable exists in the data.
2Coordination
Bring information, tools and decisions into the same context.
- 03
Confirm or correct the interpretation
Model text is not fact by itself. You correct misunderstandings and add missing context.
In the example: You correct it: “Sample size is fine; the real question is the control variable.”
- 04
Investigate the direction worth taking
The first requested solution is not assumed right; alternatives, risks and required evidence are separated.
In the example: Three paths are compared: only writing an explanation, running an additional analysis, extending the limitations section. The evidence and time each needs are visible.
- 05
Connect sources and files to the work
Where something came from and whether it was actually used stays in the same work record.
In the example: The relevant paragraph of the report, the method section of the draft and the columns of the data table are linked to the work.
3Production
Prepare the work and outputs through permitted tool paths.
- 06
Make the critical decision
Method, external action or a binding choice waits for the right person.
In the example: You decide: “Run a linear regression with the control variable”, with assumptions and limits in front of you.
- 07
Execute only authorized work
Nothing exceeds the boundary given; an unavailable tool is never presented as working.
In the example: The computation runs with the table and method you chose: a descriptive summary and the regression. If a method that cannot run today is requested, that is said plainly.
4Evidence
See sources, calculations, gaps and points needing review.
- 08
Separate action from outcome
Finishing a plan or computation does not mean the real-world effect happened.
In the example: A finished computation does not mean the reviewer is convinced. A response letter and change list are prepared as drafts; each claim shows which output supports it.
5Value
Evaluate the outcome for use and, where appropriate, the next decision.
- 09
Observe the outcome
Real measurement, documents or expert judgment update the outcome as they arrive.
In the example: When the journal's decision arrives it is added to the same work; acceptance is never promised.
- 10
Keep learning in context
The next step builds on the same evidence, decision and outcome history.
In the example: In a second revision round the same sources, method decision and reviewer objections are waiting.
What happens in real life?
Uncertainty and failure are not pushed outside the product.
What if I arrived with the wrong solution?
MMT does not treat it as fact; it first separates the goal and alternatives.
What if data is missing?
Missing evidence, acquisition options, time, risk and the decision point become visible.
What if a tool is unavailable?
The work remains; the unavailable capability is stated and no result is fabricated.
What if external action is needed?
You see the action, effect and boundary; it cannot run without separate authority.
What if the result differs?
The first proposal is not defended; new evidence updates work, judgment and next step.
What if membership stops and later resumes?
Account, work, files and history remain; new execution waits for appropriate access.
For a better future
Describe the real work as it is.
Start without choosing an industry, model or technical specialist; access opens by invitation.