This Nigerian-American tech genius created speech recognition software for African languages


Abake Adenle is a Nigerian-American entrepreneur and developer of speech recognition and speech synthesis software for African languages. At the time she started her startup, there was no known player in the speech recognition and speech synthesis software industry that focused on African languages.

Growing up, Adenle wanted to be a university professor, but she began her career in finance and worked as a quantitative strategist at Morgan Stanley for eight years. “My role was to design and develop trading strategies for institutional investors. And I’ve been doing that for about eight years, ”she told Techpoint Africa.

While at Morgan Stanley, Adenle developed an app for learning the Yuroba language called the Speak Yoruba App. According to her, she was inspired to develop the app after seeing her niece and nephew use a mobile application to learn English.

“And it did well on the App Store,” she said. “I got a lot of requests from people asking me to make a version of the app for the African language they spoke, which was nice and something I thought I could go into. But I had a job where I worked from 8 a.m. to 11 p.m. and that wasn’t realistic at the time. “

Eventually, after eight years, she quit her job in finance to focus on her speech recognition app. The move, in her opinion, was motivated by her desire to own her time and enjoy her intellectual freedom.

“I like the idea of ​​owning my time,” she told Techpoint Africa. “I also enjoy intellectual freedom. I’m working on something that I want to work on, “she emphasized,” not because I was told to work on it, but because I enjoy working on it. That’s something I appreciate. “

At the time of entering the speech recognition field in 2017, it was a relatively undeveloped area and full of uncertainty. But she was unimpressed. During the development of her app, Adenle advised on finance and also looked for financing options for her startup.

“My job was to develop speech recognition and speech synthesis as well as to implement a product with two African languages. So I made a Beta One for Yoruba and Swahili, ”she said.

“We developed a male Swahili voice and three female Yoruba voices, one of which is based on my voice speaking Yoruba. We also developed our central speech recognition model along with some variants that were content-specific speech recognition models for Yoruba and Swahili. “

It should be known that voice recognition is Siri and Alexa’s underlying technology.

While Adenle’s business proposal was laudable, many venture capitalists were unwilling to invest, so she sent a business proposal to Innovate UK and raised $ 500,000 in seed capital. “Even though we are still in private beta, we have the broadest coverage of African mother tongues to date,” she said.

The tech entrepreneur was born in the United States to Nigerian parents. However, she spent the first six years of her life in Nigeria before moving back to the United States and then returning for her secondary education.

She went on to the United States, where she graduated from Morgan State University, Maryland with an electrical engineering degree. She also pursued a Ph.D. Program in signal processing with a focus on Bayesian Inference from Cambridge University.

Today she is the founder / CEO of ajala, a London-based start-up that develops enterprise language technologies for low-resource languages ​​with a focus on African languages. Adenle sees her startup as one of her highlights. It is fulfilling for her to develop something innovative like the app she developed.

In July, Women in Voice (WIV), a US-based nonprofit, presented Adenle with the Diversity, Equity and Inclusion (DEI) in Voice Award. The award celebrates and enhances the diversity of women and genders in language technology. The Nigerian-American financial analyst was honored for her outstanding contributions to “Opening up African languages ​​for language solutions”.

Source link


Leave A Reply