Former OWASS NSMQ Champion Paul Azunre Makes AI That Lets English Speakers Speak African Languages

• Former Opoku Ware School NSMQ champion Dr Paul Azunre has demonstrated a new artificial intelligence tool developed by Khaya AI
• The technology translated an English-speaking video into Ewe while generating matching speech and lip movements
• Khaya AI already develops translation, speech recognition and text-to-speech technology for several African languages

Former Opoku Ware School National Science and Maths Quiz (NSMQ) champion Dr Paul Azunre has drawn attention to an artificial intelligence technology designed to make African languages more accessible in the digital space.

Azunre, who captained OWASS to victory in the 2002 NSMQ, shared a demonstration of technology developed at Khaya AI, where he serves as a co-founder.

His former school remains listed as the 2002 national champion, while his profile on the NSMQ website identifies him as captain of that winning team. NSMQ

Paul Azunre demonstrates AI translation tool

In the video shared on X, an original clip of Azunre speaking English was transformed into a version in which he appeared to speak Ewe.

Read also

"Goal is not to silence": Shatta Wale writes to Mahama with 7 proposals on social media abuse

According to the scientist, the demonstration combined voice translation, avatar generation, and lip-sync technology, supported by Khaya AI's next-generation text-to-speech system.

He explained that the idea was to allow a person to speak once in English and have the content localised for audiences across Africa.

The demonstration goes beyond simply placing translated text on a video.

The generated voice speaks the translated language while the visual output attempts to synchronise the speaker's mouth movements with the new audio

“Speak once in English & have it localised all across Africa,” Azunre wrote while showcasing the technology.

While the demonstration shared by Azunre involved translation from English into Ewe, Khaya AI's broader language technology is not limited to Ewe.

Khaya AI expands African language technology

Khaya AI says its platform provides machine translation, automatic speech recognition and text-to-speech services for African languages. Its translation offering currently includes languages such as Twi, Ewe, Ga, Fante, Dagbani, Yoruba, Kikuyu, Gurene, Luo, Kimeru and Kusaal. Khaya AI

Read also

Wontumi FM/TV's sports team sees another exit after Fancy Di Maria's departure

Watch the X video below:

Its text-to-speech technology has an even wider reach. The company says its models can generate speech in 32 African languages, including Asante Twi, Akuapem Twi, Ewe, Ga, Fante, Hausa, Igbo, Swahili, Wolof and Yoruba.

Azunre's journey in technology followed his exploits at OWASS. After secondary school, he studied Engineering and Economics at Swarthmore College before pursuing graduate studies at the Massachusetts Institute of Technology, where he earned advanced degrees in Computer Science and Electrical Engineering and Computer Science.

His latest demonstration adds to efforts by Khaya AI to give African languages a stronger presence in emerging AI technologies.

PRESEC produces 31 medical doctors at KNUST

Earlier, YEN.com.gh reported on PRESEC-Legon’s 31 alumni who graduated as medical doctors from KNUST in the 2025/2026 academic year.

The milestone comes in the same year the school won its record ninth NSMQ title, making 2026 a standout year for the institution and its alumni.

Source: YEN.com.gh