Together

• 

  Together Logo

Inspiration

We kept noticing the same pattern — group chats full of “we should do something soon” that never actually went anywhere. As AI becomes more embedded in daily life, it's becoming easier than ever to stay home and let an algorithm entertain you.

We wanted to build something that used AI to push back against that — not to replace human connection, but to make it easier for it to actually happen.

The spark was simple: what if your phone understood what you and your friends loved, and just… made a plan?

What it does

Together is an interest-graph-powered social app that removes the friction between wanting to do something and actually doing it.

You sign up through an AI chat that learns your interests naturally — no forms required. This builds your personal interest graph, which maps against your friends’ interests to find overlap.

The app then surfaces real events, activities, and places tailored to what you collectively love — whether you're going solo, with one friend, or organising a group.

The more you show up and attend events, the more points you earn, unlocking exclusive events and rewards.

How we built it

We built the frontend in Lovable and the backend on Supabase, which gave us authentication, a relational database, and real-time data out of the box.

The interest graph on the home screen is a hand-positioned SVG node graph that visualises a user’s interests and their friend connections. The AI onboarding chat is a guided conversational flow that maps responses to interest categories stored in Supabase.

Recommendations are ranked by interest overlap, counting how many of a user’s interests match each event’s tags.

The entire stack — from idea to working MVP — was built during a single hackathon session.

Challenges we ran into

Getting the graph to feel alive without a physics engine was harder than we expected. We initially experimented with dynamic layouts but ultimately decided to hand-position the nodes and focus our effort on interactions and hover states instead.

The onboarding flow was another challenge. Making a scripted chatbot feel genuinely conversational required more iteration than we anticipated.

We also had to make some tough scoping decisions early. Features like maps, real event scraping, and live AI calls were cut so we could ship a polished core experience instead of a half-finished full product.

Accomplishments that we're proud of

The graph view is the feature we're most proud of. When you land on the home screen and see your interests and friends radiating out from you, it immediately communicates what the app is about without requiring any onboarding.

We're also proud of the design system we built: a dark, graph-first aesthetic inspired by AI tools like NotebookLM and modern chat interfaces. It gives the product a distinctive feel compared to traditional social apps.

Most importantly, we're proud that we were able to ship something that looks and feels this complete within a hackathon timeframe.

What we learned

One of our biggest takeaways was the importance of scoping ruthlessly and designing first.

We spent time defining the visual language and interaction model before writing any code. That upfront design work made implementation much smoother because every build decision had a clear reference point.

We also learned that the most important screen in an app isn't always the most complex one. The onboarding chat is technically simple — just a few text bubbles and an input field — but it’s the moment that determines whether users trust and engage with the product.

What's next for Together

Our next step is integrating a real AI layer so the onboarding conversation becomes fully dynamic and the recommendation engine understands context — things like “I haven't seen Maya in a while” or “we usually do outdoor things on weekends.”

After that, we plan to introduce live event sourcing through venue partnerships and external APIs, along with a group planning mode that surfaces the single best activity for a specific friend group.

Long term, we see Together becoming the social operating system that connects your interest graph with the real world — helping people spend less time scrolling and more time actually doing things together.

Built With

• lovable