Inspiration
Millions of families lose their stories every day, not because they don’t matter, but because they’re never recorded. Conversations with grandparents or loved ones often feel “like something we’ll do later,” until it’s too late. We were inspired to solve this quiet but universal loss: preserving identity, memory, and culture before it disappears.
What it does
Living Legacy is a voice-first AI storytelling platform that helps families preserve memories through guided conversations. Users simply tap and talk. The system listens in real time and asks thoughtful, adaptive follow-up questions, uncovering stories that might otherwise go untold. Each session is transformed into a structured memory archive with audio recordings, transcripts, key quotes, and themes building a living, evolving biography over time.
How we built it
We built Living Legacy using a combination of real-time voice processing, LLM-powered conversational intelligence, and structured data pipelines. The frontend enables a simple, accessible “tap-and-talk” experience, while the backend uses AI (Claude) to dynamically generate context-aware follow-up questions. Conversations are transcribed, analyzed, and organized into meaningful outputs like summaries and thematic insights.
Challenges we ran into
One of our biggest challenges was making AI feel natural in deeply human conversations. We had to balance responsiveness with emotional sensitivity, ensuring follow-up questions felt thoughtful and mood-appropriate. Another challenge was handling real-time processing without lag, while maintaining accuracy in transcription and context tracking.
Accomplishments that we're proud of
We’re proud that Living Legacy doesn’t just record conversations. But rather, it transforms them into something meaningful and lasting. The system can guide conversations in a way that feels surprisingly human, uncovering stories users didn’t expect to share. We also built a fully functional end-to-end experience in a short time, from voice input to structured life narratives.
What we learned
We learned that the most impactful AI applications aren’t just technically impressive, they’re emotionally resonant. Designing for trust, privacy, and dignity is just as important as building functionality. We also learned how critical UX simplicity is when designing for users of any age.
What's next for Living Legacy
Next, we want to expand into multi-user family archives, where relatives can contribute collaboratively across time and distance. We also plan to introduce memory timelines, visual story maps, and AI-generated documentaries from collected conversations. Long-term, we envision Living Legacy as a platform that preserves not just individual stories, but entire family histories across generations.
Built With
- claude
- elevenlabs
- indexeddb
- next.js
- openai
- tailwind
- typescript
- vitest
Log in or sign up for Devpost to join the conversation.