An independent non-profit for Zomi language technology
Building the future of AI for the Zomi language
ZomiNLP develops the language data, AI models, research, and digital infrastructure the Zomi language needs to participate in the world of modern Artificial Intelligence — starting with Tedim Zomi.
Technologies we build with
- Anthropic
- OpenAI
- AWS
- Microsoft Azure
- Google Cloud
- Vercel
- Convex
- GitHub
- Linear
- Microsoft 365
- Expo
- React Native
- Next.js
- TypeScript
- Python
- PyTorch
- Hugging Face
The Mission
A place for our language in the future of AI
NLP — Natural Language Processing — is the area of AI that teaches computers to understand, translate, and work with human languages. For languages with few digital resources, that future doesn’t arrive on its own. ZomiNLP exists to build it: working across every area where modern computing can support, preserve, understand, and advance the Zomi language.
Research
Low-resource language AI is a serious scientific challenge. Our work targets publishable research at international NLP, AI, and IR venues such as SIGIR.
Technology
Real products in real hands: translation, conversation, speech, search, and writing tools for Zomi, built to professional engineering standards.
Community
Built with Zomi scholars, language experts, and native speakers — and shared with the Zomi community around the world to use, improve, and build upon.
First Product
An AI translator & assistant for Zomi
Our first major project is an AI-powered translator and conversational assistant for iOS and Android — going beyond word-for-word translation toward a system that can:
- translate naturally between Tedim Zomi and English
- understand full sentences and conversational context
- hold conversations in Zomi, like modern AI assistants
- improve over time through community participation and evaluation by native speakers
Timeline: development begins September 2026, with a first usable MVP targeted for January–May 2027.
Beyond the Translator
One mission, many projects
The translator is only the beginning. Future ZomiNLP projects may include:
Translation keyboards
Type and translate Zomi anywhere
Language datasets
Open research resources for Tedim Zomi
AI language models
Models that understand and generate Zomi
Search & retrieval
Find information written in Zomi
Speech recognition
Understand spoken Zomi
Text-to-speech
Give written Zomi a voice
Writing assistance
Grammar and writing support
Learning tools
Educational and language-learning apps
The Founder
Who’s behind ZomiNLP
ZomiNLP is founded and led by Kap Thang — Founder & President — an AI Engineer in the healthcare industry with a Master of Science in Computer Science from the University of Windsor, currently pursuing a PhD. His research in NLP and Information Retrieval includes three peer-reviewed publications, with work presented at SIGIR.
“I want to apply that experience to something much closer to home: building serious AI technology for the Zomi language and the Zomi community.”
Get Involved
Two ways to make it possible
Contribute
Zomi scholars, language experts, educators, native speakers, software engineers, and designers — help build the technology and validate the language work. Contributors are publicly credited in everything they help build, including research authorship where it is earned.
join@zominlp.orgSponsor
Sponsorship funds the tools, computing, and infrastructure — never ownership or control. ZomiNLP remains independently governed, and sponsors are clearly acknowledged across our website, products, and projects.
contact@zominlp.org