Seenit is a group-based social app where identity is foundational to context and connection. As a solo founder, I needed to design an identity system that was instant, scalable and safe, without building moderation or account infrastructure.

1. Manual uploads introduces friction and requires too many steps
2. A mix of low-quality selfies and empty states made the group-based UI feel cluttered and disconnected
3. Account management and moderation added ongoing engineering and maintenance overhead

I designed an automated identity pipeline using TikTok Login Kit to retrieve authenticated metadata. That metadata is transformed through a generative model into a stylised avatar native to the Seenit universe.

Because AI-generated identity carries risk of misrepresentation, users receive four variations and can regenerate entirely. This preserved user agency while maintaining automation.

The same generated identity persists across all surfaces of the app, from widgets and invites to reactions, without requiring profile editing or setup.

1. Reduced the time-to-first-POV from minutes to 10 seconds by removing all profile setup hurdles
2. Eliminated need for manual moderation and profile infrastructure
3. Created a reusable identity layer deployed across widgets, reactions, and invites