Nobody can tell you who you’ll fall for. We don’t pretend otherwise.
What we can do is narrower and more useful: rule out the people who were never going to work, weigh the things that actually hold a marriage together, and read both profiles properly before anyone is suggested to anyone.
Then we introduce two people and get out of the way.
Whether there’s a spark is yours to find out. You’ll see photos, a short video, and enough about someone to decide whether you want fifteen minutes with them. That part has never been something a computer could answer, and we’re not going to act like it is.
What can be known, and what can’t.
In 2017, researchers collected more than a hundred traits and preferences from speed-dating participants and applied machine learning to predict who would be drawn to whom. The models could predict how much a person tended to like others, and how much others tended to like them. What no model could predict — from anything people reported beforehand — was the spark between two particular people. The finding was replicated in 2023.
That finding shapes everything here.
On a different question — what makes a marriage last — the evidence is strong. The strongest and most replicated finding in the literature is also the least surprising one: similarity predicts stability, and complementarity does not. More than two hundred studies since the 1950s point the same way, and shared religious commitment is among the most robust versions of it — both the denomination and, more importantly, the level of observance.
That is what we weigh. Everything else, we hand to you.
A survey of 1,736 Orthodox respondents found good character the most cited reason marriages succeed — and bad character the most cited reason they end.
Three things happen, in this order.
01
First, we rule people out.
Halachic requirements and questions of status. Whatever either of you has said you will not cross. Where a match can be from. Anyone who fails here is never scored, never ranked, and never shown to you — and you are never shown to them.
And it runs both ways or not at all. You have to fit what they described and they have to fit what you described. A match that only works in one direction isn’t a match.
02
Then we weigh what’s left.
Religious life and hashkafa. Values and what you each want your life to look like. Age. Family background. How you each think about money. Education and background.
Each of these is weighted by how well the evidence actually supports it, not by how interesting it sounds. Where an answer is missing, we’re less confident about the pairing. Nothing is ever guessed to fill a gap.
03
Last, we read.
Before anyone is suggested to anyone, ShidduchAI reads both profiles in full — not just the answers, but the way each of you talked about your own life in the conversation.
That last read can hold a suggestion back. It cannot push one through.
It is also where character, and the way you each tell your story, are taken seriously — as a judgement, not a score.
What we refuse to measure.
Every one of these is available to us. Some would be easy to sell. They’re not here because they don’t survive contact with the evidence.
- Attractiveness
- We don’t score anyone’s face. Symmetry algorithms are unreliable, carry demographic bias, and attraction is exactly the thing the research shows is personal and unpredictable. You see photos and a video, and you decide for yourself.
- Astrology
- Your birth date tells us your age. Nothing else. Readings feel accurate because they’re written to feel accurate about almost anyone.
- Personality types
- Four-letter type systems have no established validity for predicting how a relationship goes.
- “Shared wounds”
- The idea that people should be matched by their childhood injuries is popular in some therapeutic writing and unsupported by evidence. We ask about your life because your story says something about your values and your resilience — not to pair you by your pain.
- Anything we can’t actually measure
- If a factor sounds impressive and we have no honest way to assess it, it isn’t in here. We’d rather do less and mean it.
What the shidduch system already got right.
Worth saying plainly, because it cuts against the usual posture of a technology company arriving in a community: on the questions where the evidence is strongest, the tradition was already there.
Shared religious life and shared values are the best-supported predictors of a marriage that lasts — and they’re the first thing anyone asks about here. Family involvement and community show up in the research as protective, not as interference. Meeting with intention, knowing why you’re being introduced, having people around you who know both families: these aren’t quaint. They’re the parts that hold up.
What the system costs is search. Finding the few people who genuinely fit, across a community larger than any one person’s circle, is slow and depends on who happens to know whom. That’s the part worth changing, and it’s the only part we’re trying to change.
The judgement stays where it belongs — with you, your family, and a shadchan if you want one.
Limits, in our own words.
No algorithm can promise a good marriage, and we won’t. What we can promise is that nothing here is guesswork dressed up as science, that we’ll tell you why we suggested someone, and that every decision is yours.
The weighting here is where the evidence points today, not a finished answer. As real introductions accumulate we’ll refine it against what actually happens in this community rather than what held in largely Western, largely secular research samples. Where something turns out not to matter, we’ll drop it.
The framework isn’t a secret — it’s published research, and it’s listed below. The implementation is ours.
References
- Joel, S., Eastwick, P. W., & Finkel, E. J. (2017). Is romantic desire predictable? Machine learning applied to initial romantic attraction. Psychological Science, 28(10), 1478–1489.
- Eastwick, P. W., Joel, S., Carswell, K. L., Molden, D. C., Finkel, E. J., & Blozis, S. A. (2023). European Journal of Personality, 37(3), 276–312.
- Joel, S., et al. (2020). Machine learning uncovers the most robust self-report predictors of relationship quality. PNAS, 117(32), 19061–19071.
- Myers, S. M. (2006). Religious homogamy and marital quality. Journal of Marriage and Family.
- Dew, J., Britt, S., & Huston, S. (2012). Examining the relationship between financial issues and divorce. Family Relations, 61(4), 615–628.
- Nishma Research (2025). Survey of the Orthodox Jewish community (N = 1,736).
When it is time, we will introduce you.
Begin your profile