Code Encoders!

Real people. Real places. Right now.

On Scene is a location-based dating app built around a simple bet: showing up beats swiping. Code Encoders designed, built, and has kept shipping the product since day one, across the mobile app, backend, and marketing site.

5M+active users
3M+live matches from check-ins
11cities live or launching
Live at a venue near you

onscenedating.com Check in, get seen, start talking with people who are actually there right now.

Client

On Scene App Inc.

Platforms

iOS & Android

Role

Full-stack build partner, design to production

Engagement

Ongoing since 2023

The Brief

Every dating app promises a match. Almost none prove someone's actually there.

The product started under a different name, Singles On Scene, before repositioning to On Scene for launch. The core idea stayed constant through that whole process: instead of matching people in the abstract, show them who's currently checked in at a bar, coffee shop, gym, or event nearby, right now.

That single idea shaped almost every technical decision that followed, from how location data gets pulled in, to how the app treats a match as something that can happen face-to-face, not just in a chat thread.

The Build

What's actually running under the hood

Code Encoders owns the full lifecycle end to end: Figma product design, engineering, production deployment, and technical documentation, on a straightforward, production-grade stack chosen to ship fast and stay maintainable as the feature list grew.

Design

Figma

Full product and UI design, from wireframes through final screens, before any code is written

Mobile app

Flutter

Single codebase across iOS and Android

Website

WordPress

Marketing site and city-by-city landing pages

Admin dashboard

React

Internal dashboard for managing users, transactions, moderation reports, push notification campaigns, blocked users, and CMS-managed profile taxonomy (ethnicity, religion, political affiliation, and more)

Backend

Node.js / Express

REST API serving the mobile app and admin tooling

Database

MongoDB Atlas

Profiles, check-ins, chat, and subscription state

Infrastructure

AWS

Application hosting plus S3 for media storage

Location data

Google Places API

Venue discovery across 8 selectable categories, pulled from Google's full 96-category place index

Payments

RevenueCat

In-app purchases via Apple Pay and Google Pay

Email

SendGrid

Verification codes and transactional messages

How It Works

The flow, in the app's own words

Five modules, built to feel like one continuous idea rather than five separate features bolted together.

Check in

Browse nearby scenes: bars, night clubs, coffee shops, kava bars, gyms, retail, event venues, restaurants, inside a default radius, either as a list or a color-coded map where pin color maps to who's there. One active check-in at a time; check out to move on. A four-hour auto-checkout keeps the data honest.

Google Places, real-time

Seen on Scene

Once checked in, a lightweight group chat opens for that venue (Feeling, Activity, and Dares), plus the option to view individual profiles and send a wink or a kiss directly, no pre-matching required.

Swipe

A conventional card-swipe layer sits alongside the location-based one, for anyone who'd rather browse from the couch. Winks, kisses, rewinds, and instant chat on a mutual match.

Chat

Matched conversations, plus controlled direct-messaging for premium users. Unmatch, delete, report, and block are all one tap away. Every report routes to a review queue without alerting the other user.

Future check-ins

A premium feature that flips the model around: instead of checking in now, book a venue and time slot in advance and see how many others already have plans there. Booking logic blocks overlapping slots and respects each venue's actual opening hours.

The Snag

Apple didn't buy "dating app" as different enough

The first App Store submission came back rejected under Guideline 4.3(b), Spam: in Apple's read, the app duplicated features already common across a saturated dating category.

Apple App Review

Your app primarily includes dating features that duplicate the content and functionality of similar apps in a saturated category. We encourage you to reconsider your app concept and submit something that offers a unique experience.

Our response

We rewrote the pitch around what actually sets On Scene apart from a standard swipe app: real-time check-ins that show who's physically present at a venue right now, and future check-ins that let users see who's already planning to be somewhere before they arrive. Both lean on shared physical presence and timing, not another matching algorithm. We updated the app's metadata, screenshots, and description to make that distinction explicit and resubmitted.

Resubmitted, approved, and live today on both the App Store and Google Play.

Still On Scene

The build didn't stop at launch

Code Encoders has stayed on as the ongoing technical partner, running weekly change cycles off direct client and field-testing feedback, most turned around in a matter of days.

28 Jul

Display names

Switched profile displays from full legal names to usernames across the app, including thumbnails and the activity feed, for privacy.

Shipped

28 Jul

Gender vs. orientation split

Separated gender (male / female / non-binary) from sexual orientation as distinct fields, rebuilding filters and matching logic around the cleaner model.

Shipped

28 Jul

Photo editor

Added in-app cropping, rotation, and zoom to the profile photo upload flow.

Shipped

28 Jul

Notification campaigns

New admin-side module for sending targeted push campaigns ("you've been near this venue for a while, check in?") via Firebase.

In progress

Caught in the Field

A client tester and a friend sat side-by-side at the same venue; one could check in, the other couldn't. The app was reading them as roughly a third of a mile apart. We traced it to GPS drift and tightened the check-in proximity threshold down to roughly a tenth of a mile, close enough to make sitting at the same bar reliably register as being on the same scene.

FAQ's

Frequently Asked Questions

On Scene is a location-based dating app that shows users who is currently checked in nearby, at a bar, gym, coffee shop, or event, rather than matching people in the abstract through swiping alone.

A Flutter mobile app for iOS and Android, a Node.js/Express backend, MongoDB Atlas for data, AWS for hosting and storage, and the Google Places API for venue discovery.

Apple's first review rejected it under Guideline 4.3(b) for duplicating features common across the dating app category. Code Encoders resubmitted with the app's metadata and description rewritten to foreground its real-time and future check-in features, and it was approved.

Over 5 million active users and over 3 million live matches from check-ins, as self-reported by On Scene on its own site.

Yes, the engagement is ongoing since 2023, with Code Encoders running weekly change cycles based on direct client and field-testing feedback.

A growth proposal covering 24 features across four pillars: product foundation, a social graph (stories, follows, discover feed), monetization, and an AI layer called Wingman for venue recommendations and conversation assistance.

Yes. Code Encoders owns the full lifecycle: Figma product design, engineering across the mobile app, WordPress marketing site, and an internal React admin dashboard, production deployment, and technical documentation. The admin dashboard itself covers user management, transactions, moderation reports, push notification campaigns, blocked users, and CMS-managed profile fields.

The Outcome

Where it stands now

5M+

Active users

3M+

Live matches from check-ins

2

App stores, iOS and Android

Figures as published on onscenedating.com, self-reported by On Scene, not independently audited.

MiamiOrlandoTampaFort LauderdaleSt. PetersburgWest Palm BeachBoca RatonLakelandAtlanta, newNew York City, newLos Angeles, new

What's Next

The next chapter is already scoped

Code Encoders put together a full growth proposal for On Scene's next phase: 24 features across four pillars, anchored by an AI layer branded Wingman.

Pillar 01

Foundation

Fixing the empty-scene cold start and tightening the core loop.

Pillar 02

Social graph

Stories, a follow system, shareable profiles, and a discover feed.

Pillar 03

Monetization

Paid reactions, venue placement, and a creator tier beyond subscriptions.

Pillar 04

Wingman AI

Venue recommendations, timing predictions, and conversation assistance.

Ready to Transform Your Business?

Let's discuss your project and explore how we can help you achieve your goals. Get started with a free consultation today.