STORYFOREST

AI-powered personalized storybooks for children. Enter your child's name and interests, and Gemini 3 creates magical bedtime stories with custom illustrations and voice narration.

Comment

Inspiration

Every parent knows the magic of bedtime stories. But finding the perfect story that matches your child's interests, uses their name, and teaches valuable lessons is nearly impossible. We wanted to create a tool that brings personalized storytelling into every home—making bedtime magical again while strengthening the bond between parents and children.

What it does

StoryForest is an AI-powered children's storybook generator. Parents simply enter their child's name, age, and interests (like dinosaurs, space, or princesses), and Gemini 3 creates a complete 10-page illustrated storybook tailored specifically for that child.

Key features:

  • Personalized Story Generation: Gemini 3 creates age-appropriate narratives featuring the child as the protagonist
  • AI-Generated Illustrations: Each page comes with custom artwork matching the story
  • Voice Narration: Text-to-speech reads the story aloud
  • Voice Cloning: Parents can record their voice and the app generates the entire audiobook in their voice—perfect for when parents are traveling
  • Story Library: Save and revisit favorite stories anytime

How we built it

  • Frontend: React + TypeScript with Capacitor for mobile deployment
  • AI Engine: Gemini 3 API for story generation and image prompt creation
  • Voice: ElevenLabs API for text-to-speech and instant voice cloning
  • Backend: Firebase (Authentication, Firestore, Cloud Functions, Storage, Hosting)
  • Image Generation: Imagen API for illustrations

Challenges we ran into

  1. Voice Cloning Optimization: ElevenLabs limits voice slots, so we implemented a system that clones the voice, generates all audio, then deletes the cloned voice—keeping only the generated audio files
  2. Audio Preloading: Initial audio playback had delays, so we built a preloading system that generates all audio before entering the reader
  3. CORS Issues: Firebase Storage required careful CORS configuration for audio playback across platforms
  4. Content Safety: Ensuring all AI-generated content is appropriate for children required careful prompt engineering

Accomplishments that we're proud of

  • Complete end-to-end personalized storybook generation in under 60 seconds
  • Voice cloning feature that lets traveling parents still "read" to their children
  • Seamless mobile experience with offline audio playback
  • Beautiful, child-friendly UI that kids can navigate themselves

What we learned

  • How to effectively prompt Gemini 3 for consistent, age-appropriate content
  • Optimizing API calls to minimize latency and costs
  • Building hybrid mobile apps with React and Capacitor
  • Managing complex async workflows (story → images → audio generation)

What's next for STORYFOREST

  • Multi-language Support: Expand beyond English and Korean
  • Interactive Stories: Let children make choices that affect the story
  • Character Continuity: Create recurring characters across multiple stories
  • Print-on-Demand: Physical book printing integration
  • Classroom Edition: Tools for teachers to create educational stories

Built With

Share this project:

Updates