A RE-X solution | Aether Platform × Secretary.io
In a complex business, the assumptions keep changing: lead times, turnaround, demand, failures. We run a simulation for each set of assumptions and compare the results, so you can make better management decisions. The rule for whether to switch to another plan when the assumptions drift is set in advance, decisions are made in the meeting, and every decision is checked against actuals. AI agents do the calculations and prepare the meeting materials.
AI Solver
Optimization and Monte Carlo simulation (thousands of trial runs) solve the plan with the variability left in. There is a plan for each scenario, and when an assumption breaks, you see what switching would gain or cost and the deadline for deciding.
Runs on Aether Platform: a secure environment where AI agents run the optimization and simulations. It can also be deployed in a closed network.
WorkWithAI
Decisions and action items from the meeting become tasks directly. A planning agent and a reviewing agent handle the calculations and the meeting pack. People watch the agents work in the browser, give instructions and approve.
Runs on Secretary.io WorkWithAI and the Aether Platform AGI Terminal. Decisions, approvals and deadlines are all on record.
Verify
Past plan versions are lined up against what actually happened to see whether the gaps lean one way. The assumptions behind the gaps are corrected with actuals and fed into the next plan.
Even after a re-plan, you can still check whether the earlier decision was right.
What the meeting decides becomes a task, the AI Solver calculates, and the agents put it into the meeting pack. Decisions are recorded, and actuals are used to correct the assumptions, which go back into the next meeting and the next calculation.
Management decisions need two loops: one that looks far ahead, early and broadly, and one that follows the facts and updates the plan. Annual and monthly PDCA (the planning loop) prepares a plan for each scenario and the decision rules in advance; for day-to-day change, OODA (the response loop) updates the plan only as far as the evidence supports and acts within the rules the meeting agreed, without waiting for the next meeting. Both loops run on the same data and simulations.
The response loop feeds the planning loop. How many cases the rules decided, what they gained or cost, and how long it took to notice a change are reviewed in the monthly meeting, and the next version corrects the assumptions and the rules. Each turn of the loop makes the way you decide better. AI prepares the meeting materials, so preparation takes less time.
Case study | Aircraft engine maintenance planning
Engine shop-visit planning, the probability that the year's maintenance spend goes over budget, spare engines versus an engine pool agreement, and when to switch plans as assumptions drift. Validated with synthetic data estimated from public information, with a demo you can step through month by month.
Examples of intended applications, not customer deployments.
Tell us about the business and the decisions that keep changing under you, and we will show you how we can help.
Opens the contact form on the RE-X corporate site (in Japanese). Please mention “AI Assisted Solver”; you are welcome to write in English.