Four businesses, costed to the minute.
Each model is a fictional company run through the CostCtrl engine. Open one to see which customers, products and sites make its profit, and which ones cost it.
Manufacturing and logistics
An Australian manufacturer and distributor. Cost to serve by customer, channel, product and hub.
Healthcare clinics
Eight Portuguese clinics. Profitability by doctor, service and clinic, with capacity by clinician.
Banking and insurance
A New Zealand bank and insurer. Risk-adjusted margin by product, branch, channel and adviser.
Retail and grocery
Fourteen New Zealand grocery stores. Margin by store, category, SKU, shelf metre and basket.
One engine, different drivers. Every model runs on the same TDABC engine. The drivers fit the business: minutes per consultation in a clinic, kilometres and drops in distribution, minutes per claim at an insurer, cases picked in a grocery DC.
Fictional companies, computed figures. The businesses and their data are invented. Every figure in the dashboards is calculated by the engine from that data, and each model's assumptions page shows how.
Goal seek, for the whole business.
An AI agent works the manufacturing model through the CostCtrl MCP server. The CFO asks for 7% net margin with no job cuts, no closures and no more than 3% of revenue lost. The agent runs ten scenarios and returns a six-lever plan. A scripted replay, with every figure computed by the engine.
The agent never types a number into the P&L. It changes inputs. The engine computes the consequences, and every figure carries the rule that produced it.
Watch the replay- FY2026 net margin
- 4.23%
- Net margin in the plan
- 7.05%
- scenarios computed
- 10
- rejected by the constraints
- 2