Yali Labs
Rwandan AI company building Kinyarwanda-focused language technology, including the Alta Tokenizer and Alta foundational model.
Last updated September 27, 2026
AI crawler reads
No AI crawler has read this profile since Oct 28, 2026.
Overview
Yali Labs is a technology company based in Kigali, Rwanda, specializing in artificial intelligence for African languages, with a particular focus on Kinyarwanda. The company was founded in 2024 by Niyibizi Schadrack, Hirwa Gael, and Murwanashyaka Philbert, who came together with the ambition of building the first Kinyarwanda foundational model entirely from scratch. Frustrated by the lack of AI tools that understood the nuances of the Kinyarwanda language and local context, the founders set out to change the landscape of AI for Kinyarwanda, combining expertise in artificial intelligence, data management, business, and software engineering. Yali Labs positions itself around inclusive AI: the belief that AI should be accessible and that tools should exist from tokenization through to foundation models capable of understanding and generating text in multiple languages, including those spoken in Africa. Its flagship work includes the Alta Tokenizer, an open-source tokenization tool designed for Kinyarwanda text that can also process languages such as English and French but performs best on Kinyarwanda, and the Alta Model, a foundational model specifically engineered for Kinyarwanda and built on transformer architecture to understand the language's unique linguistic patterns. The company describes its technology as lingual, community-driven, conversational, multimodal, and innovative. It aims to make models that can understand text, images, audio, and files, and it engages in ongoing experimentation to push the boundaries of AI research and make Rwanda a leader in African AI innovation. Yali Labs also emphasizes community: it says it is building a diverse, inclusive, and equitable community and welcomes individuals from all backgrounds to contribute to its mission of developing and deploying AI technology that benefits the people of Rwanda and the world. Yali Labs operates as a private company and participates in startup programs including the Microsoft Founders Hub and the Nvidia Inception Program, and it uses infrastructure providers such as Azure and Google Cloud. Its team is described as professionals in AI, data engineering, and business, with members spanning multiple nationalities. The company's stated mission is to develop and deploy AI technology, and it has evolved beyond its initial single-language-model mission toward democratizing AI access and enabling local talent to create solutions for specific regional needs.
History
Yali Labs was founded in 2024 in Kigali, Rwanda, by three founders: Niyibizi Schadrack, Hirwa Gael, and Murwanashyaka Philbert. According to the company's own account, the three came together with a shared ambition to build the first Kinyarwanda foundational model (FM) entirely from scratch. They were frustrated by the lack of AI tools that truly understood the nuances of the local context in the Kinyarwanda language, and they set out to change the landscape of AI for Kinyarwanda. That path led them to build custom tools to pave the way toward their goal. The founders combined expertise spanning artificial intelligence, data management, business, and software engineering. After identifying a critical gap in the AI landscape — existing language models failed to capture the intricate linguistic patterns and cultural nuances of Kinyarwanda — they were inspired to develop proprietary solutions that would serve the specific needs of Kinyarwanda speakers and set new standards for AI development in African languages. At the heart of Yali Labs' innovation is the Alta Tokenizer, a tokenization tool designed to convert Kinyarwanda text into tokens with high accuracy, increasing a model's ability to resonate with the Kinyarwanda language. The Alta Model is a foundational model specifically engineered for Kinyarwanda, built on transformer architecture to understand the language's unique linguistic patterns. The company states that it has evolved beyond its initial mission of creating a single language model. It now describes itself as standing at the forefront of a broader technological revolution, focused on democratizing AI access and enabling local talents to create solutions that address specific regional needs. Yali Labs frames its work around inclusive AI, aiming to build tools from tokenization to foundation models that can understand and generate text in multiple languages, including those spoken in Africa. As part of its growth, Yali Labs has joined startup programs including the Microsoft Founders Hub and the Nvidia Inception Program, and it relies on infrastructure providers such as Azure and Google Cloud. Its team is described as professionals in AI, data engineering, and business, with members from multiple nationalities, and it presents itself as building a diverse, inclusive, and equitable community around its mission of developing and deploying AI technology that benefits the people of Rwanda and the world.
- 2024
Yali Labs founded in Kigali, Rwanda by Niyibizi Schadrack, Hirwa Gael, and Murwanashyaka Philbert.
- 2024
Company sets out to build the first Kinyarwanda foundational model from scratch.
Alta Tokenizer developed as an open-source Kinyarwanda tokenization tool.
Alta Model developed as a Kinyarwanda foundational model on transformer architecture.
Yali Labs joins Microsoft Founders Hub and Nvidia Inception Program.
Products and positioning
Inclusive, Kinyarwanda-first AI research and language technology for Rwanda and Africa.
Alta Tokenizer
Open-source Python library for tokenizing Kinyarwanda text, based on Byte Pair Encoding (BPE); can also process English, French, or similar languages with lower compression rate; supports custom tokenizer training.
Alta Model
Foundational model specifically engineered for Kinyarwanda, built on transformer architecture to understand the language's unique linguistic patterns.
Sources5 sources
Cite this profile: Cite the canonical profile. /brand-wiki/yali-labs · Editorial policy · How profiles are compiled