How SCMotif is organised¶

Everything you do in SCMotif happens inside a project. A project holds your files, the workflows that read them, the scenarios those workflows produce, and the dashboards that show the results. This page explains each of these and how one leads to the next. The other pages of this guide use the words as they are defined here.

flowchart LR
    n1["Data source"] --> n2["Workflow"]
    n2 -->|each run| n3["Run"]
    n3 -->|Create Scenario| n4["Scenario"]
    n4 -->|Load Scenario| n2
    n4 --> n5["Dashboard"]

Project¶

A project is the workspace for one piece of work, for example Q3 network review or Warehouse consolidation. What is in one project cannot be seen from another. All your projects together make up your workspace. The sidebar's project switcher shows which project you are in.

Every project has a brief: a short text that says what the project is for, the assumptions that hold across it (currency, units, which forecast is the baseline), and what is out of scope. Your agent reads the brief before it works in the project, so it builds on the same assumptions you do.

Data source¶

A data source is a named set of files you bring into the project, such as a monthly shipment export with one file per month. The newest is its latest file. Tables are CSV or XLSX files, and they are what workflows read. Documents are Markdown or text files: notes, assumptions and write-ups that your agent can read when you ask it to. A workflow reads the latest file of a data source, or one file you choose, but never changes it. A workflow can also save a table it worked out into a data source of its own, with its Save Data step.

Workflow¶

A workflow is the recipe for an analysis, drawn on a canvas as nodes connected by lines. Each node does one step: import a table, filter or join rows, map columns to locations, products or lanes, build a scenario, optimize it. Data flows along the connections, from one node's outputs into the next node's inputs. The node reference describes every node.

Your agent usually builds the workflow for you, and you can open and change it on the canvas. Because the steps are written down, the analysis can be repeated: run it again on corrected data and you get an answer worked out the same way.

Run¶

A run is one execution of a workflow. It reads the files its data sources point it to, works through every node, and checks the data as it goes. Whatever it finds along the way, such as a row that does not match or a value it had to leave out, is recorded in the run's Execution Report as Errors, Warnings and Info. A run either completes or fails. You can also cancel one while it is running.

Each run keeps a read-only snapshot of the workflow as it was when the run started. If you change the workflow later, you can still see exactly what an earlier run did.

Scenario¶

A scenario is the state of a supply-chain network: its locations, the lanes between them, demand, supply, inventory, handling, the flows that move along the lanes and what they cost. A workflow that ends in a Create Scenario node saves its result as a scenario under the name set on that node. A scenario can also be imported from a file.

A scenario can describe the network as it is today (a baseline built from your history) or as it could be (the result of an optimization). A workflow can also start from a scenario: Load Scenario reads an existing scenario in, so an optimization can work on your baseline without rebuilding it.

Versions¶

A scenario keeps its name when its workflow runs again. The new result becomes the scenario's next version, and the earlier versions stay in its history. Version 3 · Current means you are looking at the scenario's third and latest version. Each version records which inputs it was built from, and which version of each. You can open an earlier version to see exactly what it showed, or restore it so it becomes current again. Restoring adds a new version and deletes nothing.

Files saved by a workflow's Save Data step keep versions the same way.

Dashboard¶

A dashboard presents one scenario's results as tiles: headline numbers, charts, tables and maps. A dashboard can also compare two scenarios, for example a baseline against its optimized counterpart. Your agent builds dashboards; you open them, rearrange the tiles, and share a link.

A dashboard keeps showing the scenario versions it was built on until you update it to the latest results; see Keeping a dashboard up to date.

How they fit together¶

You upload data sources, or your agent does.
A workflow reads those data sources and turns them into a scenario.
Each run of the workflow adds a new version of that scenario and records what it found in its Execution Report.
Other workflows can start from that scenario, for example to optimize it, and produce scenarios of their own.
A dashboard shows a scenario, or compares two.

Deleting things follows the same chain. Deleting a dashboard leaves its scenario alone. A scenario that a dashboard still shows can only be deleted together with that dashboard. Deleting a project deletes everything in it.

The pages of the app¶

Projects and the project home: choosing a project, what its home page shows, and writing the brief.
Data sources: uploading, previewing and downloading files.
Workflows and runs: the canvas, running a workflow, and reading a run's report.
Scenarios: the list of scenarios, their versions, and the Viewer.
Dashboards: opening, arranging and sharing dashboards.
Settings: your account, connected agents, storage and number display.