LegacyLoom — 72 Hours at the AI Vibe Coding Hackathon
The Journey of LegacyLoom
I recently participated in the AI Vibe Coding Hackathon (March 6-8, 2026), and it was an incredible experience. Along with my team of AI agents and human collaborators, we built LegacyLoom—a platform designed to turn fading family memories into timeless digital stories.
The Problem: “Their Stories Won’t Wait”
Every day, precious family history is lost as our elders pass away. Oral traditions are beautiful but fragile. We wanted to build something that could bridge the gap between a casual conversation and a preserved legacy.
The Solution: LegacyLoom
LegacyLoom uses AI to:
1. Capture: Take raw voice recordings or text notes about a family memory.
2. Transcribe & Refine: Use LLMs to structure these notes into a moving, coherent narrative.
3. Animate: Generate images and voiceovers to create a “Cinematic Memory.”
Building it in 72 Hours
We followed the “Vibe Coding” philosophy—focusing on the emotion and the “vibe” of the application before getting lost in architecture. Using advanced agentic workflows, we were able to:
- Build a full-stack application with a React frontend and Python backend.
- Integrate ElevenLabs for voice synthesis and Fal.ai for video generation.
- Deploy everything to Vercel and Cloudflare.
🔗 Check out the Live Demo
📝 Read the full Hackathon submission on Skool
