A price increase, a reorganization, a launch: before you announce it, Kapari shows you who goes along, who has doubts, who pushes back, and why. You fix it while there is still time.
Prefer video? Getting started, in three and a half minutesA few sentences, in your own words, the way you would explain it to a partner.
AI-simulated voices, built to resemble your customers, employees and partners; no one is interviewed. Several dozen, depending on the decision. Each one answers in its own words, with its own reasons.
Who goes along, who has doubts, who pushes back. What holds each one back (cost, workload, doubt about execution) and what they would do. Then Kapari's conclusion: what to fix before you announce.
Kapari puts the announcement in front of the people it touches: leaders, local staff, remote hires, managers, recruiters. Here is one of those voices, exactly as it came out of the test.
"They hired me fully remote, promised it in writing, and now I'm supposed to uproot my life for three days a week, no way, I'm already updating my resume."
The company learns before the email goes out that the fight is about a broken promise, not the commute. See the full test.
A few minutes.
From what you describe. Kapari reads your decision and your situation, works out which groups are affected, and builds voices that are plausible in your company, grounded in public data. Before you launch, you review them: you can remove some, add some, or add a whole group.
Across the 100 decisions in our exam, each run three times, 87 reached the same conclusion all three times.
ChatGPT writes well. Kapari runs every version of your decision past the same voices and gives you a conclusion you can defend. The full comparison.
It is not a poll and it is not a prediction: Kapari shows you the possible reactions and the reasons behind them, and the decision stays yours. And it has been put to the test on real decisions from the past: see the proof.
Your decisions are stored in France, visible only to you, and Kapari never uses them to train an AI. Who processes what: the Security page.
How Kapari picks the voices. It reads your text and works out who the decision affects. It draws on 67 situation types prepared in advance (return to office, layoffs, pricing and subscriptions, banking fees, union relations, higher education, among others). Each voice has a profile and a documented reason to react, for example the fear of losing what you already have, or the sense that a promise was broken. Behind them sit more than 840 verified sources (Census and Bureau of Labor Statistics surveys, public datasets, published research), each with its link. If your topic falls outside these situations, Kapari fetches public sources live.
Who writes what. The AI writes what each voice says. The sorting of groups (for, divided, against) and the conclusion come from a fixed rule, not from the mood of the AI. The reception is placed on a scale, from outright rejection to clear support.
How much each group counts. By default, every group counts the same: adding voices that agree with you does not change the result. You can give one group more weight, and that choice stays on display.
After the test. A voice surprised you? You can ask it your own questions.
We do not promise you will make the best decision. We do promise that when you make it, you will have the strongest set of elements to help you.
Kapari measures its audience with Google Analytics, to know how many people read these pages. Nothing is stored on your device until you accept.