Nigerian Startup Intron Develops Voice AI for African Accents to Tackle Healthcare Efficiency

Nigerian AI startup Intron is addressing a critical efficiency gap in African healthcare by developing specialized voice recognition technology. A 2022 study by Intron revealed that Nigerian doctors have an average typing speed of only 22 words per minute, significantly hindering their ability to document patient consultations efficiently, especially given the high patient loads in the country.

This challenge led Intron's founder, Tobi Olatunji, a medical doctor and former Amazon ML scientist, to pivot from electronic medical records to building foundational voice AI. The team discovered that existing global voice recognition systems struggle with diverse African accents, Nigerian English, Pidgin, and code-switching between languages like English and Yoruba, making them impractical for local use cases.

Intron's solution, Sahara voice AI models, is designed to accurately transcribe speech patterns specific to its target markets. This proprietary infrastructure layer can be integrated into various products across sectors, including healthcare for clinical notes, legal for court proceedings, and financial services for call centers, thereby improving operational efficiency across the continent.

The development of these localized AI models is crucial for Africa, where a shortage of healthcare professionals means doctors are often overwhelmed. By enabling faster, more accurate documentation through voice, Intron's technology has the potential to free up doctors' time, allowing them to focus more on patient care and significantly improve healthcare service delivery.