Unison is live, and the beta is free.
I'm building a different way to meet people in Telegram: rate photos, upload yours, and let the algorithm consider both people's visual preferences.
No swiping. No public profile feed.
If Unison cannot find a suitable introduction nearby today, you may have to wait. Location and the visual preferences of both people matter. I would rather explain the wait up front than set an expectation of a match every day.
After a good first date, "I had fun" is easy to read as politeness. "I had fun. Want to do it again next week?" gives the other person a real question to answer.
A good dating photo can give someone an easy opening. A picture from a hike makes "Where was that?" a much easier first message to send than "Tell me about yourself."
"Are you here to chat, or are you hoping to find a relationship?"
A useful question if the answer matters to you. It gives you a way to explain what you want, too.
Matching people by visual preferences does not replace that conversation.
One product question I keep coming back to with Unison: where should the recommendation end? For me, it ends at the introduction. Whether to start a conversation is the first decision the two people get to make.
"Thanks for talking. I don't want to take this further."
That is a complete answer. You can be considerate without turning a decision about dating into a debate you have to win.
For a weekday first date, "somewhere central" can still mean two awkward journeys home. Pick a place both people can get to easily, then see if the time works.
When a dating service promises a "perfect match," I want to know what it has actually checked.
In Unison, the model considers visual preferences. It does not know whether you share plans for the future or enjoy spending time together.
If you want a short first date, put an end time in the invitation: "Coffee from 3 to 4 on Sunday?" It is easier to say yes when both people know what they are agreeing to.
"Looking for something serious" can mean different things to different people.
"What would a relationship look like for you right now?" gives you more to talk about.
Shared words are a starting point. The details still need a conversation.
You enjoy talking to someone, but you don't feel like meeting them yet.
Do you usually need more time for interest to develop, or does that already tell you enough?
I'm building a dating service. I can explain how we choose an introduction; I can't promise a relationship.
Unison considers both people's visual preferences. A conversation and a meeting tell you things the model cannot.
If you don't want another date, you can say that without writing a review of the person.
"Thanks for meeting up. I don't want to take this further, but I wish you well."
Clear and considerate is enough to aim for.
What would you hand over to AI in dating?
With Unison, it is finding possible introductions using both people's visual preferences.
Whether you enjoy the conversation, want to meet, or want to see each other again is still yours to decide.
A dating photo can look great and still leave you unsure who you'll be meeting.
I'd rather use a recent photo with a clear view of my face.
How do you decide which photos to put on a dating profile?
"Lower your standards after 35."
Which standards? Finding someone attractive? Enjoying their company? Wanting similar things?
If someone has given you that advice, what did they actually suggest changing?
If you returned to dating after a long break, what was hardest to start again?
Writing a profile, sending the first message, arranging a date, or deciding you were ready?
I'd like to understand where a service can make that first step easier.
"We should meet sometime" leaves a lot to work out.
"Want to get coffee on Saturday afternoon? We can pick somewhere convenient for both of us" gives the conversation a place to go.
What kind of first invitation do you prefer?