Hey HN, I'm Tugay, a solo developer from Turkey. I built ImagineIf because I always wanted a collaborative version of "Choose Your Own Adventure" books — but with AI-generated visuals. How it works:
- Someone starts a chain with "Imagine if..."
- Others add 280-character segments
- AI (FLUX-dev via Replicate) generates a visual for each segment
- Stories can branch — readers pick which direction to follow
Tech stack: React Native/Expo frontend, FastAPI backend, MariaDB, FLUX-dev for image generation, Groq API for text processing. Hosted on Vercel (frontend) + custom server (backend). Supports 22 languages with automatic translation.
What I'm struggling with: Getting the first real users beyond friends and family. The platform has content but needs more active contributors to make the collaborative aspect work. Classic chicken-and-egg problem.
What's interesting technically: Maintaining visual consistency across story segments is hard. Each new image needs to match the established visual style of the chain while reflecting new narrative developments. I built a prompt-building system that carries forward visual context from previous segments.
I'd love honest feedback — what works, what doesn't, what feels off. Try starting a chain or contributing to an existing one.
Hey HN, I'm Tugay, a solo developer from Turkey. I built ImagineIf because I always wanted a collaborative version of "Choose Your Own Adventure" books — but with AI-generated visuals. How it works:
- Someone starts a chain with "Imagine if..." - Others add 280-character segments - AI (FLUX-dev via Replicate) generates a visual for each segment - Stories can branch — readers pick which direction to follow
Tech stack: React Native/Expo frontend, FastAPI backend, MariaDB, FLUX-dev for image generation, Groq API for text processing. Hosted on Vercel (frontend) + custom server (backend). Supports 22 languages with automatic translation.
What I'm struggling with: Getting the first real users beyond friends and family. The platform has content but needs more active contributors to make the collaborative aspect work. Classic chicken-and-egg problem.
What's interesting technically: Maintaining visual consistency across story segments is hard. Each new image needs to match the established visual style of the chain while reflecting new narrative developments. I built a prompt-building system that carries forward visual context from previous segments.
I'd love honest feedback — what works, what doesn't, what feels off. Try starting a chain or contributing to an existing one.