Zooni
Zooni — A creator social media platform where users capture video content in a whole new way via both the front and rear-facing cameras at once.
Long story short
Becoming the next leading creator platform with a focus on innovation and safety is no small feat. Our team delivered a sophisticated, entirely customized package for Zooni — from dual camera recording to a recommendation engine, on-device content moderation, AR filters, effects and editing tools.
Zooni Highlights
— Saw a steady 50% conversion rate just weeks after going live
— Launched with influencers incl. Piper Rochelle (10M followers)
— Broke into Top 200 Apps in the App Store (1 week post-launch)
— The average user spends 9+ min. on Zooni
Services / Deliverables
— Product Management
— Backend Engineering
— iOS Engineering
— Data Science
— QA
Elsewhere
About Zooni
For Zooni users, sharing both sides of a story gets literal. By simultaneously capturing video between the front and rear-facing cameras, users remain in the moment even as it's being shared. Video can be recorded in splitscreen, miniscreen and fullscreen. No limits.
Everything is collaborative and interactive, with cutting-edge AI at its core. Follows and comments are a given, but Zooni also offers 200 million music tracks, endless filters and editing tools, plus an engaging feed thanks to an advanced recommendation solution. And to keep the community safe, the platform uses tailored on-device content moderation.
Challenge
A team with formidable drive but no technical background, Zooni needed a tech partner that could handle the complexity and scale of their ambitions while providing guidance along the way.
The discover video feed had to recommend content perfectly suited to every user. Because the app is targeted mainly at kids, any inappropriate content had to be flagged at record speed, using technology that’s typically only accessible to giants like Apple — but without breaking the bank. And as Zooni’s main value proposition, the split camera had to function flawlessly.
Solution
To propel Zooni past any comparable service, we brought on 15 experts, including three data science engineers, to build a production-ready iOS mobile app with integrated, custom AI.
By building and implementing on-device content moderation — the only viable way to moderate video chat and live streaming — we provided Zooni with a typically inaccessible tool. Rooted in state-of-the-art AI, our solution is secure, fast and scalable, with a level of performance that was impossible just a few months prior for anyone other than mammoth companies.
For the recommendation engine, personalization was key. We created two machine learning models: one recognizes activities (what is being done in the video?), the other recognizes entities (which objects are in the video?). This allows Zooni to show highly tailored content to every user. The platform also recommends content that a user’s friends are interested in.
STRV was able to build an exceptionally safe platform that’s fun to use and doesn’t burden users with irrelevant content. It’s a true showcase of what data science can accomplish — and we’re happy to say it’s only the beginning of our long-term partnership with Zooni.
