A community-driven research collective focused on NLP and low-resource technologies for the languages of Meghalaya, including Khasi and Garo.
Bridging the gap between academic research and practical deployment for indigenous languages.
Targeted initiatives to digitize and empower Meghalayan languages.
Developing Neural Machine Translation (NMT) systems specifically for Khasi, Garo, and related dialects.
Language documentation, corpus creation, and building robust annotation pipelines.
Creating NLP systems designed for public good and real-world utility.
Building open tools for easier data collection and community annotation.
We build, curate, and release assets on Hugging Face.
High-quality data for training and evaluation.
State-of-the-art architectures adapted for our needs.
Interactive ways to test our technology.
Tynrai currently operates as a community organization. Roles and access are managed by the org admins. As the project evolves, governance structures may be formalized to ensure long-term sustainability.
Unless otherwise specified, our content is released under permissive open-source licenses (Apache-2.0). We believe in open science while respecting data sovereignty.