The AI That Hears What Google Misses: How Intron Health Is Decoding the African Voice
HealthTech, Assistive Tech

The AI That Hears What Google Misses: How Intron Health Is Decoding the African Voice

MadeInAfrica Team

Maker

Dr. Tobi Olatunji

Known For

Founding Intron Health and building the "African Medical Speech" dataset.

Tools & Equipment

Natural Language Processing (NLP), Deep Learning, Python, Electronic Health Records (EHR).

Geography

West Africa
🌍Nigeria

Coming Soon on YouTube

Why does AI struggle with African accents? Meet the doctor who built the world's largest dataset of African medical speech.

Medical documentation is slow, but for African doctors, Western AI makes it impossible by failing to recognise local accents. Tobi Olatunji’s Intron Health is fixing the 'Acoustic Gap.' By training AI on over 200 distinct African dialects, they are saving doctors seven hours a week and ensuring that no patient’s story is lost in translation.

Doctors in Africa spend up to 40% of their time on paperwork. While Western doctors use dictation software, those tools fail miserably when faced with a Nigerian, Ghanaian, or Kenyan accent. Dr. Tobi Olatunji decided to "make" the data that was missing. He built Intron Health, which has now collected thousands of hours of clinical speech from across the continent. His AI now allows surgeons in hospitals from Douala to Nairobi to dictate notes with 99% accuracy, freeing them to save more lives.

Lessons for Budding Makers:

1.**Data is a Product:** If an AI doesn't work for your people, the solution isn't to wait for Google, it's to build the dataset yourself. 2. **Focus on Friction:** Tobi didn't build a new hospital; he removed the paperwork friction for the doctors already there.

More Stories

Continue exploring the makers shaping Africa's future