What a matching model actually does
It compares answers. You describe values, everyday life, pace and priorities; someone else does the same; the model looks for where you are close and where you differ.
In Aivelyn, confirmed answers are compared across ten compatibility areas, and each person distributes 100 priority points so that both people's priorities weigh equally in a shared score.
What a score means - and does not mean
A high score means your answers point in the same direction in the areas you both consider important. It is decision support ahead of a meeting, not a prediction about the relationship.
Chemistry, timing, life situation and how you handle difficulty do not show up in answers. That is why the explanation behind a match is more useful than the number: it tells you what to talk about first.
Where AI matchmaking goes wrong
Three recurring pitfalls are worth knowing about.
- Answers describe how you want to be, not how you actually behave in a relationship.
- Similarity gets overrated. Agreeing on everything is neither necessary nor always good.
- An opaque model cannot be questioned - with no explanation, you cannot judge the suggestion.
What you still decide yourself
Whether to meet, how you talk to each other, what you share and when you say no - that is still yours. A model can reduce the number of people you need to consider and give you something to start from in a first conversation.
If you want to see what an explanation looks like in practice, there are examples on the match report page.