Artificial intelligence (AI) is rapidly transforming various sectors worldwide, but for many Africans, language and cultural barriers still limit access to these advancements. To solve this challenge, EqualyzAI, an Afro-centric AI startup is on a mission to ensure AI can understand and communicate in African languages, starting from Nigeria’s indigenous languages.
Using hyperlocal multimodal datasets, created in collaboration with native language speakers, the firm is building agentic AI solutions capable of understanding, processing, and generating responses in many African languages.
Speaking at a recent event marking International Mother Language Day, Olubayo Adekanmbi, founder of EqualyzAI emphasised the importance of language in preserving cultural identity.
“Everything that we represent is codified in our language,” he noted. “Several languages in Nigeria have gone extinct because they are not digitised. There is no way for the next generation to learn them.”
He highlighted that Africa is home to over 2,000 languages, yet most AI models struggle to understand a fraction of them.
International Mother Language Day, established by UNESCO in 1999, commemorates the sacrifices of the people of Bangladesh, who fought to have Bengali recognised as an official language on 21 February 1952. The event highlights the critical role of language in human development, a cause that aligns with EqualyzAI’s objectives.
Read also: The future of work in the age of artificial intelligence
Adekanmbi expressed concerns about the digital divide, where dominant languages on the internet shape AI’s learning while local languages remain marginalised.
“If today AI is learning only from languages on the internet, how many of our languages are on the internet that AI can learn from?” he asked. “If we do not increase our language presence online, we will enter another level of digital divide, a form of neo-colonialism where dominant languages dictate the future of AI.”
To counter this, EqualyzAI is training AI models to learn and process indigenous languages just as a child learns a language. “We just learn those languages, the same way a child learns a language, and we make AI learn it,” Adekanmbi explained.
The firm collects data in the form of audio, text, images, and videos from native speakers with its DataCollect solution, ensuring AI can understand context, intonation, and meaning as naturally as possible.
The company’s AI solutions aim to serve practical needs. For example, a Hausa-speaking farmer can use AI-powered tools to receive tailored farming advice based on weather forecasts and soil conditions. Similarly, a street vendor can use AI to help her track finances and plan savings.
“Imagine the number of people in Africa who can be empowered if they can just speak to their phones and get responses in their language,” Adekanmbi stated.
EqualyzAI is also developing Afro-SLM (Small Language Model), a specialised AI model that efficiently processes African languages without requiring vast computational resources. Unlike existing AI models, which struggle with code-mixed languages, where English is blended with local dialects, Afro-SLM is designed to understand and generate responses that reflect real-life language usage in Nigeria.
Adekanmbi added, “This is not just about technology; it is about empowerment and inclusion. At EqualyzAI, we are bridging this gap by creating AI solutions trained on African realities, ensuring that no one is left behind.”
By enabling AI to speak and reason in Nigeria’s indigenous languages, EqualyzAI is preserving linguistic heritage and unlocking economic and educational opportunities for millions.
English (US) ·