I spent two years building the social app I wanted to use

Built around feelings, not followers.

Last Tuesday I stepped out of a coffee shop and felt something small but real. Not big. Just the kind of something you forget to notice. I reached for my phone the way you do when you want to put a moment somewhere.

I almost opened Instagram out of habit. I caught myself, opened Atlas instead, and posted a quick check-in. Happy. Grateful. “The kind of day that doesn’t need a reason.” Friends only. No follower count to perform for. No public scoreboard. A friend loved the post within the hour.

That moment is the whole reason I built Atlas. Not the part where I shared. The part where I almost didn’t.

The slot machine in your pocket

I grew up alongside the social Internet. I have receipts on every shift it went through, from MySpace bulletin boards to Twitter quote-tweets to Instagram’s algorithmic feed to TikTok’s For You page. I watched apps I genuinely loved as a teenager become engagement-optimized machines I started avoiding as an adult.

Here is what nobody told me until I had been on these apps for over a decade. Every metric a social app shows you is a decision. Follower counts are not a feature of friendship. Public engagement rankings are not a feature of human connection. They are dopamine knobs, and the companies that build them turn the knobs to make you stay longer.

I was tired of feeling worse after I scrolled. I was tired of the version of myself that wanted to post for the wrong reasons. I was tired of friends I actually love getting flattened into a number in someone else’s growth chart.

So I built the alternative I wanted to use.

What Atlas does differently

Atlas is a complete social platform. Feed, friends, messaging, communities, places, photos. Not a journaling app. Not a wellness app. A social network. The difference is what’s at the center.

The first thing Atlas asks you is how you are actually feeling. Not “what’s on your mind.” Not “share an update.” Just a quick emotional check-in: pick from a grid of feelings, select an intensity, write a line if you want. That check-in is the post. The feeling is the point. Everything else is supporting cast.

From there the rest of the app behaves like a social network. You have friends, not followers. Friendships are bilateral: you both agreed. You can mark a subset as Close Friends, and your Close visibility tier is for them specifically. Your feed shows what your friends are feeling, with light boosts for emotional resonance. When you’ve been stressed lately, the feed surfaces posts from other people who are also stressed, so you remember you’re not the only one. When you check in happy, the feed leans toward shared happiness.

You can send a friend a vibe. There are 33 of them, ranging from a sunshine to a hug to “really?”. Small gestures that don’t demand a response, just an acknowledgment. You can join a Circle or many Circles, which are anonymous communities around a shared interest. You can check in at a venue and let your friends know you’re at your favorite coffee shop.

Here is what’s missing on purpose. No follower or following numbers on a profile. No “trending” surface. No engagement-driven feed that pushes you past your friends’ content into the slop. No public scoreboard. No concept of a “public account” that strangers can amass and turn into an audience. The graph is friends. Reactions on a post are small gestures between friends or the Atlas community, not points in a popularity contest.

The math underneath

I’m a solo technical founder, and the parts I’m most proud of are the ones nobody sees.

The feed ranking is a hybrid score. Each post in your feed gets a number: recency contributes the most (a fresh post starts strong and decays exponentially with a defined half-life), emotional resonance with your latest check-in adds a boost, your relationship with the author (Close Friend, friend, community) adds a smaller boost, location proximity adds another. The whole feed sorts on that score. No engagement-driven amplification, no recommended-from-strangers slot, no influencer pollution.

If your latest check-in was a negative emotion, the math shifts. The emotional-match bonus doubles, and a guarantee kicks in: at least two friends who are also feeling something similar appear in your top ten posts, if they exist. The product encodes the mission directly. When you are struggling, Atlas tries harder to show you that you are not alone. When you are happy, the feed just stays fresh.

Circles auto-scale so that members never see instance numbers, even when a public Circle exceeds its size limit. The system routes new joins to available capacity invisibly. From the inside it feels like a small room, always.

I filed a patent on 11 inventions covering the feed math, the visibility model, the friend tier propagation, the vibe system, the auto-scaling Circles, and a few other mechanisms. Patent Pending isn’t marketing; it’s a statement that the design decisions are load-bearing.

One person, two years, ten thousand decisions

I have not raised any money. I am one person. I wrote the iOS app, the Android app, the database schema, the moderation queue, the messaging stack, the venue system, the emotion taxonomy, this essay, and more.

There are things I have not gotten to yet. Atlas does not have stories. It does not have a creator economy. It does not have a friend-of-friend recommendation engine, on purpose. It does not have any of the things that companies in this space spend their second year on, because I’m trying to build the version that doesn’t need those things.

I am pretty sure that’s the version of social I want. I am pretty sure other people want it too. The way to find out is to put it in front of you.

What I want you to do

Try it. The app is free in the App Store and on Google Play. Sign up, check in once, invite three friends. The whole point of Atlas is that it gets meaningful at three friends. One friend feels like nothing. Three friends and it is suddenly a small private network where you can be a person.

If you don’t like it, tell me why. I read everything. I’m building this as a human with a vision of what social media should be. The things you want it to be will probably end up shaping what comes next.

If you do like it, the thing that matters most for any new social network is whether you tell someone. Word of mouth is the only growth model that doesn’t drag the product back toward the things I tried to leave out. So if it lands, share it the way you’d recommend a coffee shop, not the way you’d boost a post.