Clinical speech-recognition startup Intron Health has secured a $1.6 million pre-seed funding round to expand its specialized voice recognition tool tailored for African accents. The round was led by Microtraction, with participation from Plug and Play Ventures, Jaza Rift Ventures, Octopus Ventures, Africa Health Ventures, OpenseedVC, Pi Campus, Alumni Angel, BakerBridge Capital, and several angel investors.
The core technology addresses a significant gap in voice recognition, where speakers of minority languages or those with distinct accents often face poor performance from standard speech-recognition tools used for transcription and automation.
African Clinical Data and Technology
Intron Health, founded by CEO Tobi Olatunji (a former medical doctor in Nigeria), claims to possess Africa’s largest clinical speech database. Key statistics about its training data include:
-
Audio Clips: 3.5 million
-
Hours of Audio: 16,000 hours
-
Contributors: Over 18,000, primarily healthcare practitioners.
-
Geographic Coverage: 29 countries and 288 accents.
Drawing most contributors from the healthcare sector ensures accurate transcription of complex medical terms. The startup is confident about deploying its model in markets like Ghana, Uganda, and South Africa, where its data collection is rapidly expanding.
Founder’s Motivation and Problem-Solving
Olatunji’s motivation stems from his firsthand experience practicing medicine in Nigeria, where he witnessed system inefficiencies, particularly the time wasted on repetitive paperwork (e.g., manually writing patient names and details on every lab order).
This frustration led him to pursue Master’s degrees in medical informatics from the University of San Francisco and computer science at Georgia Tech. He then gained experience as a machine learning scientist at companies like Enlitic and Amazon Web Services, specializing in natural language processing (NLP) for healthcare.
Intron Health initially launched in 2020 with the aim of digitizing hospital operations in Africa via an Electronic Medical Record (EMR) system. However, adoption was difficult because physicians preferred writing over typing. When attempting to integrate existing third-party speech-to-text tools to automate data entry, Olatunji discovered that “thick African accents and the pronunciation of complicated medical terms and names made the adoption of existing foreign transcription tools impractical,” leading to the genesis of Intron Health’s specialized speech-recognition technology.
Impact and Future Plans
The speech-recognition tool, which integrates with existing EMRs, has been adopted in 30 hospitals across five markets, including Kenya and Nigeria. The impact has been significant; in one instance, Intron Health helped reduce the waiting time for radiology results at one of West Africa’s largest hospitals from 48 hours to 20 minutes. Such efficiencies are vital, as Africa continues to face one of the lowest doctor-to-patient ratios globally.
With the new funding, Intron Health is focused on several technical advancements:
-
Perfecting noise cancellation.
-
Ensuring the platform works well in low bandwidth environments.
-
Enabling the transcription of multi-speaker conversations.
-
Integrating text-to-speech capabilities.
The long-term plan is to add intelligence systems or decision support tools for tasks like prescription or lab test recommendations, which can help reduce doctor errors and speed up patient care.
Intron Health is also contributing to research in Africa, partnering with Google Research, the Bill & Melinda Gates Foundation, and Digital Square at PATH. This collaboration aims to evaluate popular large language models (LLMs) such as OpenAI’s GPT-4o, Google’s Gemini, and Anthropic’s Claude across 15 countries to identify potential risks of bias and ensure culturally attuned models are available for African clinics and hospitals. This work aligns with the global speech- and voice-recognition market, which is projected to reach $84.97 billion by 2032, growing at a CAGR of 23.7% from 2024.
