Moliyili (/mo-lee-YEE-lee/)

powered by nasari

House of Knowledge

Moliyili was a pre-colonial center of learning in the Kingdom of Dagbon (present-day Northern Ghana), established in the 1720s. It served as a beacon of scholarship and knowledge preservation for generations.

Today, we continue that legacy—conducting foundational research into NLP for local West African languages. Our work encompasses quality data provision, digitization of these languages, creation of foundational models, and support for language technology advancement. We are dedicated to preserving and promoting underrepresented languages for future generations.

Foundational Models

We are developing foundational language models tailored for West African languages—built from the ground up to understand the unique phonetic, tonal, and grammatical structures of these underrepresented languages.

Text-to-Speech (TTS)

Coming Soon

Natural voice synthesis for West African languages. Our TTS models are trained to capture the tonal nuances, rhythmic patterns, and phonetic subtleties essential for authentic speech generation in these languages.

Development in progress. Contact us for early access opportunities.

For Researchers

Access high-quality annotated datasets and foundational research specifically prepared for academic NLP and machine learning work on West African languages. Our collections include:

  • Parallel text corpora for translation studies
  • Phonetically annotated speech recordings
  • Morphologically tagged text collections
  • Grammatical structure annotations

All resources are documented with comprehensive metadata and follow established linguistic annotation standards. Suitable for computational linguistics research, low-resource language studies, and cross-linguistic analysis.

For Commercial Use

Licensed datasets and foundational language models available for integration into commercial language products and AI applications, including:

  • Voice assistant and speech recognition systems
  • Machine translation platforms
  • Language learning applications
  • Content localization services

Our commercial resources come with flexible licensing options and ongoing support. We work with organizations to ensure proper implementation and optimal performance of West African language features in production environments.

Inquiries

For access to datasets, collaboration opportunities, or general inquiries about our work, please reach out to us.

Email: saabiqsaha@gmail.com

Contact: LinkedIn Profile