StoryBridge is an AI-powered prototype built for an AI & Healthcare hackathon that helps preserve seniors’ life stories and turn them into engaging, illustrated narratives for younger generations.
The project focuses on bridging the emotional and generational gap between seniors and their families by transforming real conversations into accessible, visual stories that children can understand and connect with.
StoryBridge takes a recorded conversation between a senior and a caregiver and converts it into a short comic-style story by:
- Transcribing spoken audio from real conversations
- Cleaning and summarizing the story into a clear first-person narrative
- Breaking the narrative into individual scenes
- Generating a comic-style image for each scene
The final output is a simple, human-centered story that helps grandchildren see their grandparents as people with rich experiences, not just patients.
- Preserves personal and cultural history
- Encourages empathy across generations
- Supports meaningful engagement in caregiving contexts
- Has applications in elder care and memory care settings
- Demonstrates responsible, human-centered use of AI
- Python
- Deepgram (Nova-3) — conversational speech-to-text
- Moonshot AI (Kimi-K2-Instruct) — narrative summarization
- FLUX.1-dev (Black Forest Labs) — comic-style image generation
- Hugging Face Router — unified access to AI models
- Python 3.9 or higher
- API keys for:
- Deepgram
- Hugging Face
- OpenAI-compatible router
pip install openai huggingface_hub deepgram-sdk
- End-to-end AI pipeline design
- Speech, text, and image model integration
- Prompt engineering for narrative control
- API orchestration across multiple AI providers
- Human-centered problem solving in healthcare contexts
- Rapid prototyping under hackathon constraints
StoryBridge was developed as a proof of concept during an AI & Healthcare hackathon, with a focus on accessibility, ethical AI usage, real-world healthcare relevance, and emotional impact.
- Web-based interface (Next.js)
- Live transcription with caregiver prompt suggestions
- Multilingual translation support
- PDF / eBook export for families
- Secure storage for private stories
Galit Tauber + Jisara Wuttikreekiat