Children's Story Generation App (LLM → TTS)

Situation

Creating children's stories on demand is a time-consuming creative task — most parents don't have the time or inspiration to invent them, and generic generation lacks structure and a natural reading experience.

Task

Build an app that generates an original children's story from a short conversational prompt and reads it aloud with natural speech, delivered through a mobile app.

Action

I'm building the product end-to-end:

• Agentic generation: a LangGraph flow manages the interaction — gathering the child's or parent's input, structuring the story, and keeping the tone age-appropriate.
• Text-to-speech: the story is rendered to audio with ElevenLabs voices.
• Front end & backend: a Next.js app with Firebase for state and delivery to mobile.

Result (in progress)

The end-to-end pipeline — conversation → original story → narrated audio in a mobile app — is built and close to publication. Final results pending release.

Stack highlights

• LangGraph for the interactive generation flow
• ElevenLabs for text-to-speech
• Next.js for the app front end
• Firebase for backend state