Shaili Sanchaya: Reviving Indic Typography from Archival Metadata

While the digital ecosystem for Latin scripts thrives with thousands of open-source typefaces, Indic typography - particularly for languages such as Kannada - remains severely underrepresented. Most contemporary Indic fonts are designed for modern publishing and do not preserve the historical letterforms, ligatures, and typographic variations found in nineteenth- and early twentieth-century print traditions.

At Sanchaya (Sanchaya.org), we are exploring how the extensive digital archives created through the Servants of Knowledge initiative can help revive early print heritage. We have developed Shaili Sanchaya, an open-source toolchain for revival font development that extracts historical glyphs and letterforms from scanned public-domain books and pre-computer-era print collections to create modern, standards-compliant Unicode typefaces.

Beyond preserving typographic heritage, revival fonts are essential for improving OCR accuracy on historical documents. By digitally recreating legacy typefaces, we enable better recognition of older publications and help unlock vast collections of Indic knowledge that remain inaccessible to search, research, and dissemination.

We will present our open-source workflow - from archival scanning and character extraction to glyph normalization, vectorization, and font engineering—and discuss the challenges of revival font development, as well as how such efforts can lay the foundation for large-scale Indic glyph datasets and future machine learning applications.

• The Indic Typography Gap: The lay man’s understanding the design and technical constraints of digital Indic typefaces.

• Archival to Digital Pipeline: How to leverage community-digitized public archives (like Servants of Knowledge) for historical font revival.

• FOSS Font Engineering: Talk about open-source tools? or tell the story of what help was out there to start thinking about unicode-compliant Indic fonts.

• The Bigger Picture: How opening up glyph datasets can revolutionize OCR, archival indexing, and accessibility across regional Indic languages