Guide
Getting Started
SSPLAX models a constrained decision as a Network, Composition, Portfolio, or Assignment. It finds the best feasible answer for the selected objective, then shows which limits shape it.
The core loop
A typical SSPLAX session follows this pattern:
Pick a starter model, describe the decision to the AI builder, or start blank. Its Form defines whether you are routing, blending, selecting, or assigning.
Choose what to optimize: maximize throughput, minimize cost, or minimize resource consumption. Or flip the question and seek a required target with the least intervention.
Set capacities, budgets, and rates, by hand or from a spreadsheet. Mark which limits are protected and which are negotiable, and name the scenarios worth planning against.
The solver returns the answer in the same Form you built, with the achieved outcome, governing limits, supporting details, and a verified repair package when no plan is feasible.
Stage a quick what-if, pressure-test the limits, check whether committed plans survive your scenarios, and build a decision brief from what holds up.
Your first model
To get familiar with the workspace, open an example and change one input.
- 1. Go to Examples and open any model. Clinical Supply Coverage is a good place to start.
- 2. Click Open this model. You'll land in the workspace with the template loaded.
- 3. Pick an objective and proceed to the Constraints screen.
- 4. Move a constraint slider, tightening a budget or lowering a capacity limit, then select Solve.
- 5. Results opens on Answer. Use Analysis to understand it, Compare to test a separate scenario, and Report when you are ready to share it.
You've now run a constrained optimization, found the active limits, and seen how the recommendation shifts when an assumption changes. The rest of the guide takes each step further.
Key concepts
These terms appear throughout SSPLAX. For full definitions, see the Glossary.
The shape of the decision: Network, Composition, Portfolio, or Assignment. It controls the authoring and Results grammar, not the mathematics the solver is allowed to use.
A stage in your system: a source, service, queue, or sink. Nodes are connected by flows.
A path between two nodes. Each flow carries one product with min/max bounds, cost, yield, and optional resource consumption.
A limit on the system: a budget cap, capacity ceiling, minimum delivery, or resource limit.
What you want to optimize: maximize throughput, minimize cost, or minimize resource use.
An explicit choice the solver can take or leave, such as a supplier to qualify or a shift to add. Grouped with the options it competes against, and carrying its own costs and effects.
A named alternative future with its own limits and rates. Scenarios can be weighted and planned against as a set.
Committed choices are locked in before you know which scenario happens. Adaptive ones can change once you do.
A constraint fully used up in the baseline answer, known in solver terms as a binding constraint. Relaxing it can improve the answer.
How much the objective improves when you relax a limit by one unit, also called its shadow price. A high value flags a limit worth a closer look.
How much room is left before a limit becomes active, also called slack. Plenty of headroom usually means the limit is not driving the decision.