It was a dismissal. A signal that language, in that moment, was power, and access to it was not freely given. For Owoade Apotierioluwa, that moment did more than sting.

It defined a problem he would spend years trying to solve. Saki sits close to the Benin Republic, where French is widely spoken. In theory, that should have made learning the language easier.

In practice, it exposed a quiet divide. Apotierioluwa did not have a stable French teacher through junior secondary school. By the time he reached his senior classes, he was struggling, scoring as low as 4 out of 20.

It stood in sharp contrast to his performance in other subjects. But the real challenge was not just academic. In the boarding house, students from Francophone backgrounds would switch from English to French when he walked into the room.

Conversations would continue, but without him. It created a strange kind of isolation, being physically present, yet excluded. “It felt like we were strangers in our own home,” he recalls.

Determined to close the gap, he tried to learn from classmates. When that failed, he improvised. He began exchanging provisions for lessons, building informal partnerships just to keep up.

Slowly, he improved, eventually finishing near the top of his class. Saki, Oyo State, Nigeria What could have been a setback became a decision point. He chose to study French at university.

At university, the pattern repeated itself. The language barrier was no longer personal. It was systemic. Many students struggled, not because they lacked intelligence, but because they lacked foundational exposure to the language and its cultural context, which hindered their ability to engage with the material effectively.

This was where something shifted. Apotierioluwa stopped seeing French as a subject to pass and started seeing it as a problem to solve. He began using tools like Google Lens to translate text and support his own learning.

Later, he introduced classmates to early AI tools, helping them prepare for exams and navigate coursework. Also read: PewBeam: All you should know about Dara Sobaloju’s scripture rendering AI app The spark of RedApt, scaling the vision The idea that would later become RedApt was quietly forming through lived experience. The formal spark came later, triggered by frustration.

While trying to make his pastor’s messages accessible to a wider audience, Apotierioluwa turned to existing AI voice and translation tools. They fell short, struggling with code-switching, incantations and speaking in tongues. African accents, like the specific inflections of Nigerian Pidgin, or tonal variations.

Context was lost. The output felt disconnected from the original message. What if your Pastor Could speak Portuguese?When we started building https://t.co/WNzHQNk00W, it was because we had tried using Elevenlabs and it couldn't recognize the Nigerian accent and would sometimes mix words up when we tried to use it for church messages.But no more pic.twitter.com/g5oJIEHqUe— Eri aka Cucaracha? (@ApotieriO) March 31, 2026 That was the breaking point.

But it was not the beginning. “All my life, I have been looking to close that language gap,” he says. “Because of the alienating factor of not being able to speak a particular language.” Out of that realisation came RedApt, an African AI startup focused on dubbing and translation.

Its goal is straightforward but ambitious: to help creators translate audio and video content into over 50 languages, with a strong focus on African nuance. The emphasis on nuance is deliberate. Many existing systems struggle with African accents, tonal variations, and cultural context.

When those elements are lost, meaning is distorted. RedApt is built to address that gap. The company did not emerge from a well-funded lab or a network of elite connections.

It started, as Apotierioluwa puts it, “in the ocean, in a wooden boat, without food or water”. The founding team came together through the church, Winners Chapel International. Comprising Maryann Nnaji, David Mac-asore, and Emmanuel Ibiang, each member brought a distinct skill set, from machine learning to product development.

Before RedApt, some had worked together on a branding agency that generated modest revenue. But building an AI product at scale was an entirely different challenge. There were no investors.

No cloud credits. No roadmap. At one point, the team was using their salaries to pay for hosting, leaving little for basic needs.

Progress was slow. For nearly two years, attempts to build the product yielded limited results. Then came a turning point. After applying unsuccessfully to a cloud support programme, Apotierioluwa received an unexpected follow-up t