Intron Health
9 known investors
Intron provides speech-to-text and text-to-speech APIs designed for African accents and local languages, powering voice applications such as clinical ambient listening, medical dictation, judicial transcription, voice bots, and call center automation. It positions itself as a voice interaction infrastructure layer for use cases across Africa.
Also known as Intron · Intron Health
Investors · 9
Also in the syndicate · 5
Funding
SEC filings, press & company announcements$1.6M disclosed across 1 of 3 rounds · 2024
- $1.6Mpre-seedJul 2024 · 4 sources
Microtraction (lead), Africa Health Ventures, Alumni Angel, Baker Bridge Capital, Jaza Rift Ventures, Octopus Ventures, OpenseedVC, Pi Campus, Plug and Play Ventures
Source ↗
Source: company announcements and press reports — follow each round's link for the claim.
Company profile
researched Aug 2026Intron Health is a voice AI company that builds automatic speech recognition (ASR) and text-to-speech models tuned for African accents, languages and natural code-switching. Its clinical speech recognition platform supports more than 200 accents spoken in developing countries, starting in Africa, and is used to dictate and transcribe medical content directly into electronic medical records; the company reports up to 92% accuracy on medical terminology spoken with heavy accents and roughly a sevenfold reduction in documentation time. The speech product was originally built on top of an EMR that Intron supplies to hospitals, though more customers use the speech-to-text product than the EMR itself. Features include clinical entity extraction (diagnoses, laboratory tests, medications, procedures), administrator-facing quality control and text analytics over clinical notes, configurable retention periods for patient data (for example 24 hours, one week or one month), offline/on-premise deployment for settings without reliable internet, and browser-based access from any device.
Beyond healthcare, the company packages its models into product lines spanning fintech, customer experience, justice, and surgical use cases, plus platform components for data, voicebots, form autofill and offline deployment. Its current model release, Sahara v2.5, adds bilingual language-mixing support across 12 African languages (extending a Swahili-English capability from v2), a first trilingual model covering English, Kinyarwanda and French, text-to-speech with code switching in 13 languages, and seven additional languages that take total coverage from 56 to 63 languages, alongside general recognition-quality improvements. Listed supported languages include Afrikaans, Amharic, English, Hausa, Igbo, Kinyarwanda, Luganda, Oromo, Pidgin, Shona, Swahili, Wolof, Yoruba and Zulu.
Deployments cited by the company include Branch International collections voice agents, Audere's Self-Cav reproductive health WhatsApp chatbot in South Africa, ARM Investments (transcription and summarisation), the Ogun State Judiciary for court transcription, a Nairobi health centre for clinical note generation, and offline medical scribe use at PAMO Clinics in Nigeria. The company also publishes research (including a Bio-RAMP Lab) and has announced an upcoming 2026 Africa Voice AI Report.
Founding story
Founder and CEO Tobi Olatunji is a physician who practised in Nigeria, working at one of West Africa's largest hospitals where he routinely saw more than 30 patients a day, and later worked as a machine learning scientist at companies including Amazon Web Services and Enlitic on clinical natural language processing. Intron was launched in 2020 to digitise healthcare, and Olatunji quickly identified data entry as the main bottleneck to electronic medical record adoption: heavy patient traffic meant thousands of keystrokes and up to six hours a day on paperwork for some doctors, making digitisation impractical for already overworked clinicians. That insight led the company to build a clinical speech recognition platform for African accents.
Business model
Intron sells speech recognition and text-to-speech technology to institutional customers — hospitals and clinics, financial services firms, call centres and government judiciary bodies — as a platform accessible through browsers and devices, with both cloud-native and on-premise/offline deployment options. It also provides an EMR system into which its dictation and transcription tools integrate, and pursues integration into other major EMRs.
Sources describe institutional deployments of the speech platform and EMR to hospitals and enterprises but do not specify pricing or contract structure.
Traction
As of the July 2024 pre-seed announcement, Intron Health served more than 30 public and private hospitals — including Aminu Kano Teaching Hospital, Asiwaju Bola Ahmed Tinubu Hospital in Lagos, Babcock Teaching Hospital in Ogun and Meridian Health Group in Nairobi — reaching more than 56,000 patients, with doctors in Nigeria, Ghana, Kenya, South Africa and Uganda completing documentation seven times faster. Company-reported customer outcomes include over ₦1.2 million in delinquent loans recovered in one week by Branch International collections agents built on Sahara, growth from 3 to 9 courts at the Ogun State Judiciary over about a year, conversion of a 14-minute Swahili-English consultation in Nairobi into a structured clinical note in under 30 seconds, and over 2,000 consultations with a 4x documentation speed improvement at PAMO Clinics using an offline medical scribe deployment.
Latest developments
The current release is Sahara v2.5, which introduces bilingual language-mixing across 12 African languages, a first trilingual model spanning English, Kinyarwanda and French, code-switching text-to-speech in 13 languages, seven newly added languages for a total of 63, and improved recognition quality for real-world African speech. The company is also promoting an upcoming 2026 Africa Voice AI Report.
▸Full profile — market position, technology, go-to-market, geography, history, risks & controversies
Market position
The company describes itself as having built Africa's first clinical speech recognition platform and the first bilingual language-mixing support across 12 African languages, positioning itself against global assistants and general-purpose ASR trained mainly on North American and European accents. Coverage notes that with over 3,000 of the world's estimated 7,000 languages and dialects in Africa, many African languages and accents remain excluded from mainstream speech systems. StartupList Africa groups it with Ubenwa Health and MyHealth Africa as comparable companies.
Differentiation rests on accent and language coverage (200+ African accents; complex African names), medical terminology handling in heavily accented speech, code-switching and language-mixing support, offline and on-premise deployment for low-connectivity environments, direct deployment into EMR workstations rather than mobile copy-paste workflows, customer-controlled data retention and a stated policy of not training models on customer patient data, plus administrator-facing clinical quality control and entity extraction. Independent Medical Question Answering benchmarks conducted for the Gates Foundation and CLEAR Global by Gooey.AI covered 4 of 6 Nigerian languages.
Technology
The core technology is a family of speech models (Sahara, currently v2.5) trained on African speech, including bilingual and trilingual code-switching models, ASR with accent and dialect handling, and text-to-speech with code switching across 13 languages, with 63 languages supported in total. Training rests on a proprietary clinical speech dataset the company describes as Africa's largest: over 3.5 million audio clips across multiple specialties and domains, covering 288 accents from more than 18,000 contributors in 29 countries, which allows deployment in new hospitals with minimal additional fine-tuning. The stack supports both cloud-native and on-premise/offline operation, real-time transcription, structured clinical note generation, and clinical entity extraction.
Go-to-market
Distribution runs through direct institutional deployments and pilots that expand over time — for example a state judiciary engagement that grew from 3 pilot courts to 9 with a target of all 18 courts — alongside partnerships with local healthcare providers, a strategic partnership with NVIDIA tied to a Commonwealth AI initiative, and content marketing via a blog, research publications and a planned Africa Voice AI report. Proceeds from its pre-seed round were earmarked partly for expanding distribution.
Public and private hospitals and clinics and their clinicians, financial services companies (lending collections, asset management), customer-experience and call-centre operations, judicial and legal institutions, surgical teams, and health-focused NGOs deploying voice chatbots.
Geography
Headquartered in Lagos, Nigeria, with clinical deployments reported in Nigeria, Ghana, Kenya, South Africa and Uganda, judiciary work in Ogun State, Nigeria, and health chatbot voice support in South Africa; the training dataset draws contributors from 29 countries.
History
Founded in 2020 as a healthcare digitisation effort, the company shifted focus to clinical speech recognition, building its platform on top of the EMR it supplies to hospitals; speech-to-text adoption subsequently outpaced EMR adoption. It assembled a large proprietary clinical speech dataset, partnered with NVIDIA as a strategic partner for a Commonwealth initiative promoting AI use in Commonwealth states, and in July 2024 announced a $1.6 million pre-seed round led by Microtraction to deepen research, strengthen cloud-native and on-premise capabilities, expand distribution and hire. It has since broadened from clinical documentation into fintech collections, customer experience, justice and surgical products, and released the Sahara model line, with v2.5 adding language mixing, trilingual support and expanded text-to-speech.
Risks & controversies
Reported operating constraints include unreliable internet connectivity in target markets, which the company addresses with offline deployment, and the handling of sensitive patient data, for which it states it does not train models on customer patient data and offers configurable data retention periods.
Compiled by commissioned research from 8 cited public sources — announcements, filings, and press listed under research sources below.
Key figures
latest reportedCompany-reported or press-reported figures, each dated to when it was claimed — not independently audited.
Timeline · 6
launches, deals, and filingsOver about one year of work with Intron, deployment grew from 3 pilot courts to 9 courts, with the stated goal of covering all 18 courts in the state.
Sahara-powered collections agents built with Branch International recovered more than ₦1.2 million in delinquent loans in one week, including after-hours and weekend repayments on long-delinquent loans.
Sahara v2.5 introduces bilingual language-mixing support across 12 African languages, a first trilingual model (English, Kinyarwanda, French), text-to-speech with code switching in 13 languages, seven new languages taking coverage from 56 to 63, and improved recognition quality.
Intron Health teamed up with NVIDIA as a strategic partner for a Commonwealth initiative to promote the use of AI in Commonwealth states, alongside its effort to integrate its software into major EMRs.
The platform is used by doctors across Nigeria, Ghana, Kenya, South Africa and most recently Uganda.
Pre-seed round of $1.6 million led by Microtraction with participation from Plug and Play Ventures, Jaza Rift Ventures, Octopus Ventures, Africa Health Ventures, OpenseedVC, Pi Campus, Alumni Angel and Baker Bridge Capital, plus angels from Google, CLEAR Global, NYU and Optum. Proceeds allocated to research, cloud-native and on-prem capabilities, distribution expansion and hiring.
$1.6M source ↗
Dated company events from announcements, filings, and press; legal rows summarize public dockets and regulator releases.
In the news
▸Research sources · 8
primary sources listed
- Intron Healthintron.io · web
8 public sources were cited for this profile; the first-party ones are listed here.
Frequently asked questions
- What does Intron Health do?
- African-accent speech recognition and text-to-speech company whose Sahara models power clinical, court, call-centre and voicebot workflows.
- Who are Intron Health's investors?
- Intron Health's investors include Microtraction, Octopus Ventures, Pi Campus, Plug and Play Ventures.
- How much funding has Intron Health raised?
- Intron Health has disclosed $1.6M raised across 1 of its 3 known rounds.