A Costa Rican folkloric dance company had decades of performances, costumes, photographs, and cultural documentation with no centralized system to preserve or access them. Materials were stored in boxes, hard drives, and personal collections, at risk of being lost.
The Solution
I built an AI-powered digital archive that preserves the company's cultural heritage and makes it accessible to researchers, members, and the public.
How it works:
Physical and digital materials are uploaded through a simple intake interface
Gemini analyzes each item, extracting descriptions, dates, people, events, and cultural context
Items are automatically categorized by type (performance, costume, document, photo, video)
The archive supports full-text search, tag-based browsing, and timeline navigation
Public and private access tiers control what's visible to different audiences
Key features:
AI-powered metadata extraction and cataloging via Gemini
Multi-format support (photos, videos, documents, audio recordings)
Timeline view showing the company's history chronologically
Costume and performance catalogs with detailed descriptions
Public portal for community access and education
Admin tools for curators to review and enrich AI-generated metadata
Tech Stack
AI: Gemini
Backend: Python
Database: Supabase
Results
2,000+ items digitized and cataloged in the first month
Archive accessible online for the first time in the company's 30+ year history
Used by 3 university research projects within the first quarter
Preservation of at-risk materials that had no digital backup