Artificial Intelligence is acquiring the ability to converse in African tongues, courtesy of these ventures
In the dynamic landscape of artificial intelligence (AI), Nigerian startups Intron and Spitch AI are making significant strides in leveraging voice AI technology to address local needs and improve service delivery across various sectors.
Intron, founded by medical doctor Tobi Olatunji, is a trailblazer in using voice-based AI to help healthcare professionals enter medical records using speech. The startup's clinical speech recognition tools have been instrumental in clearing documentation backlogs, particularly in hospitals across Nigeria and Kenya. Intron's technology understands medical terminology, allowing doctors to dictate notes hands-free in real-time, a feature that is particularly beneficial in busy hospital environments [1][2][4].
The startup's AI models are also being used in courtrooms, such as the Ogun State Judiciary in Nigeria, to transcribe court proceedings in real-time. This innovation reduces hearing times and allows judges to focus on deliberations rather than manual note-taking [1][3]. Furthermore, Intron offers voice AI solutions for call centers and media, providing human-like conversational agents that enhance customer support and media inclusivity [3][4].
Spitch AI, another startup building voice AI tuned to the African continent's linguistic realities, is working to bridge gaps in areas such as customer support in indigenous languages and making services more inclusive by developing voice-based AI solutions tailored to local needs [4]. Spitch AI offers simple APIs and SDKs that let any team plug local-language voice capabilities into call centers, media tools, or learning platforms. Developers buy pay-as-you-go credits to use Spitch AI's speech-to-text and text-to-speech APIs, while large corporations pay bespoke fees for tailored models [4].
Starting with Yoruba, Spitch AI has since added Hausa, Igbo, Nigerian-accented English, and Amharic to its repertoire [4]. The company also offers a no-code conversational agent builder and supports early users in various industries [4].
Intron's operations involve managing massive linguistic diversity and paying thousands of contributors for annotated voice data. The startup manages over 3,000 active contributors for this purpose [1]. Intron builds its AI models from scratch, requiring real research talent and a significant investment of time and resources. Training a single model for Intron takes weeks of GPU time and cloud orchestration, and costs can be high [1].
Meanwhile, the Nigerian government is also investing in AI development, commissioning Awarri, a Lagos-based startup, to build the country's first LLM model trained on five low-resource Nigerian languages and accented English, in partnership with Data.org [1]. Awarri launched a data annotation lab and LangEasy, a platform that enables anyone with a smartphone to contribute to training the model through voice and text inputs [1].
These advancements in voice AI technology are set to revolutionize sectors such as agriculture, media, and finance by enhancing communication and information dissemination in indigenous languages and accents, potentially improving outreach and engagement. By addressing the underrepresentation of African data in global AI models and developing solutions tailored to local needs, these startups are contributing to a more inclusive and efficient digital future for Africa.
This report was produced with support from the Centre for Journalism Innovation and Development (CJID) and Luminate.
[1] Source: The Guardian Nigeria [2] Source: TechCrunch [3] Source: TechCabal [4] Source: Ventures Africa
- The fintech sector may see innovation with the integration of voice AI technology, as both Intron and Spitch AI are working to improve customer support in indigenous languages.
- In the logistics sector, startups like Intron are developing voice AI solutions that aim to streamline documentation processes, such as clearing medical records backlogs in hospitals and enhancing efficiency in busy environments.
- In the realm of startups and lifestyle, the development and adoption of voice AI technology are contributing to a more inclusive and efficient digital future by addressing the underrepresentation of African data in global AI models and catering to local needs.
