We built an AI matchmaker that optimizes for relationships, not swipes

Hey PH 👋

I’m the CTO at LoveActually. I’ve spent most of my career building recommendation systems, and dating apps have always bothered me from an engineering standpoint.

Most of them are optimized for engagement: swipe more, match more, keep you coming back. But very few are built to help two people actually build something that lasts.

That frustration is why we built LoveActually.

What we kept hearing from users:

• “I spend hours swiping and still meet the wrong people”

• “Profiles don’t show who someone really is”

• “We match, then the conversation just dies”

Our takeaway was simple: it’s not a UX problem — it’s a signal quality problem.

Static profiles are basically dead data. Swiping is noisy. And most matching systems more or less stop caring once a match happens.

So we asked a different question: what if matching was based on how people actually communicate and behave over time — not just photos and tags?

How LoveActually works differently under the hood:

1. Dynamic personality modeling
   People aren’t fixed profiles. The way you write, interact, and make choices updates an evolving personality model over time.

2. Outcome-driven matching
   We focus on long-term relationship outcomes, not swipe rates or short-term reply rates.

3. Scenario compatibility (not checklist matching)
   Two people can look perfect “on paper” and still fall apart in real life. We try to evaluate fit across real situations — weekends, social energy, planning style, emotional pace, etc.

4. Support after the match
   Matching is step one. We put a lot of effort into what happens after: helping conversations keep momentum, making communication less awkward, and taking friction out of planning a first date.

One thing that surprised me building this:
The models weren’t the hardest part.

The hardest part was resisting vanity metrics. It’s incredibly easy to make a dating app feel “busy.” It’s much harder to make it feel meaningful.

We intentionally chose slower, higher-confidence matches — even when it hurt our short-term numbers.

Why I’m posting here:
LoveActually isn’t about “more matches.”
It’s about fewer, better ones.

If any of this resonates — using tech for something real instead of engagement loops, building for long-term outcomes, or just thinking most dating apps are technically lazy — I’d love to hear your thoughts.

Happy to answer any technical or product questions.

Thanks for reading ❤️
— CTO, LoveActually