In a time when technological advancements heavily drive global language adoption and preservation, Africa’s rich and diverse linguistic heritage has often been sidelined. While we’ve been seeing a surge in AI-powered language tools worldwide, not much thought has been paid to African languages until now when a beacon of hope has emerged through IgboSpeech, an innovative AI model created by Igbo-American Ijemma Onwuzulike.

Preserving a language is one thing; making it readily accessible to all and sundry is another. Launched on July 1, 2024, IgboSpeech is an advanced speech-to-text AI model that automatically converts Igbo audio to text. This Automatic Speech Recognition (ASR) model is specifically designed for the Igbo language, and it’s the first-ever speech-to-text model that caters to the needs of over 30 million people worldwide who speak the over 20 known Igbo dialects.

This is Onwuzulike’s second language tech invention—after Nkọwa okwu— that’s bolstering the preservation and promotion of the Igbo culture. Nkọwa okwu is an Igbo language learning tool that has already made significant strides in bringing the richness of the Igbo vocabulary to a broader audience. The e-learning platform currently has the largest free, open-contribution, multidialectal, and audio-supported Igbo-English dictionary, featuring over 8,000 words, 3,000 Igbo sentences, and more than 1,700 word audio recordings.

While Nkọwa okwu is ideal for beginners, the newly launched IgboSpeech is designed for those who already speak the language, offering valuable support for both personal and professional use.

To hear it in Onwuzulike’s words, “It supports people who already speak Igbo.

“This would be incredibly helpful for translators, people contracted to translate things from Igbo to English or to write out large bodies of text.”

Since its launch earlier this month, IgboSPeech has attracted over 1,900 visitors from Nigeria, the United States, the United Kingdom, and Canada.

Onwuzulike’s exploits in language-linked tech are deeply rooted in her educational background. She studied Computer Science and Japanese Language and Literature at Dartmouth College, Hanover, USA, from 2015 to 2019. Currently employed as a software engineer at Google, she continues to push the boundaries of technology.

After creating the first Igbo speech-to-text AI model, Onwuzulike has hinted at developing a text-to-voice and translation model, with plans to apply for the Lacuna Fund’s Natural Language Processing (NLP) grant to support this endeavour.

“I’m considering applying again for the NLP grant to build Nigerian language audio-text datasets. Reach out if you want to work on a project together,” Onwuuzulike tweeted in June.

Being a Lacuna Fund grant finalist played a pivotal role in the emergence of Nkọwa okwu over two years ago. Securing the grant this time is bound to propel Onwuzulike’s latest innovations even further.

Though being the founder and CEO places Onwuzulike at the centre of this cultural resurgence—from Nkọwa okwu to IgboSpeech—the impact of her invention wouldn’t be felt this far and wide without the commitment of her team and a host of volunteers across various tech fields who are dedicated to this cause.