Archival Intelligence
New Orleans' cultural archives — newspapers, sheet music, personal letters, photographs, and early recordings documenting the birth of jazz — are deteriorating in under-resourced institutions. Many exist only in fragile physical form. Without intervention, this irreplaceable record of one of America's most significant cultural contributions will be permanently lost.
Archival Intelligence is a Schmidt Sciences HAVI-funded research project developing AI techniques to preserve and analyze these endangered archives. Led by Katherine Elkins (PI) and Jon Chun (Co-PI) at Kenyon College, the project was one of 23 teams selected worldwide from a highly competitive international pool. The work is conducted through the Human-Centered AI Lab.
Photo: Louisiana State Museum.
Our approach
The project develops AI pipelines combining computer vision, OCR, and multimodal analysis to process archival materials that resist conventional digitization. View full research →
Multimodal Archival Analysis
Computer vision and OCR optimized for degraded historical documents — newspapers, sheet music, letters, and photographs that resist standard digitization.
Data Sovereignty
Communities whose heritage is being preserved maintain control over how their archives are accessed, represented, and used. Preservation as partnership, not extraction.
Preventing Cultural Flattening
When physical archives disappear, the cultural perspectives they contain are erased from AI training data. Preservation is both a cultural imperative and an AI equity issue.
Recent highlights
Could AI Save Endangered Archives?
In-depth feature on the team's first meeting, the challenges facing archives nationwide, and how a smartphone-based AI tool could change everything for small museums and communities.
Kenyon Receives Major Gift for AI Work
The official announcement of Kenyon's $330,000 award from Schmidt Sciences to develop a free, open-access AI system for rescuing endangered archives.
Kenyon College Receives $330K Gift for AI-Humanities Project
Local coverage highlighting how this globally recognized project originates from rural Knox County, Ohio — and its potential impact on archives around the world.
Collaborating universities
Our interdisciplinary team spans four institutions, bringing together expertise in computer science, jazz studies, digital media, and the humanities.
Louisiana State University
Supratik Mukhopadhyay contributes AI and machine learning expertise in multimodal data processing, pattern recognition, and scalable computational systems.
Columbia University
Brent Hayes Edwards brings deep knowledge of jazz history, African American cultural production, and the complexities of cultural archives.
Berklee College of Music
Lori Landay provides insight into how musical and cultural archives can be meaningfully preserved while maintaining lived context and cultural significance.
Get involved
We're building an open, collaborative network. Whether you work with archives, study AI, or care about cultural preservation — there's a place for you.
Partner with Us
Do you manage or work with an endangered archive? We're looking for partner collections — especially those with multilingual, multimodal, or underrepresented materials.
Join Our Network
Jazz historians, musicologists, archival scientists, AI researchers, digital humanists — join our growing network of consultants and contributors.
Research Opportunities
Kenyon students and alumni are at the heart of this work. Contact us about research opportunities, or share how your experience connects to our mission.