Shina Arogundade founded MyItura in January 2023 to build a digital health platform that could make patient records interoperable across Nigeria's fragmented healthcare system. Three years on, AI sits at the centre of every product the company has launched, from electronic medical records to lending. Arogundade says the company would have struggled to win over providers without it.
MyItura is one of the startups interviewed for the AI Made in Africa project, a Briter and GIZ-commissioned study that tracks how startups on the continent adopt artificial intelligence and navigate infrastructure constraints, gathering evidence to help the African Union and key partners shape future policy.
MyItura started with a simple electronic medical records (EMR) platform. The company expanded to build a telemedicine platform, a laboratory test booking platform, and a "care now, pay later" service called MediLoan. According to Arogundade, AI was necessary for the company's next stage of growth. "Probably without AI, we wouldn't even be able to provide that holistic support to the providers or would have struggled more with adoption from a lot of the providers," he noted.
During its early development phase, getting patient records was difficult because providers were reluctant to release them to other hospitals. MyItura introduced telemedicine to bridge this gap, but doctors found typing up consultation notes to be a highly tedious task, with many preferring to write them by hand. To solve this, MyItura deployed an AI tool that transcribes doctor and patient conversations and summarises them into standard medical notes for future reference.
Beyond transcribing telemedicine consultations, AI is embedded across MyItura’s entire suite of products. The startup uses optical character recognition (OCR) within its electronic medical records platform to help hospitals instantly digitise paper records through a simple scan, bypassing the need for manual data entry. MyItura’s MediLoan lending product also uses machine learning trained on internal data and anonymised third-party lender data to offer instant credit responses for medical emergencies. Lastly, for the startup’s test booking platform, an AI tool automatically interprets complex laboratory results in formats patients can read without clinical knowledge. can easily understand, helping them make informed health decisions.
Navigating unstructured data, talent, and cloud costs
The first constraint, Arogundade says, is data. Training on foreign datasets produces hallucinations; local data is too unstructured to substitute. “For instance, the way one hospital or even a single doctor captures patient data can differ from another, and separate laboratories often use different reference numbers for the same medical tests,” notes Arogundade. To overcome this data gap, MyItura is actively working to build its own datasets rather than relying solely on foreign data wrappers. To combat hallucinations, they put their systems through continuous, rigorous training and extensive reviews.
Another challenge is talent acquisition and salary competition. Early in the company's life, MyItura hired a developer with a master's in machine learning to build out its early systems. Within months, a foreign employer offered her $9,000 a month. "There is no way I can match that," Arogundade says. "As the founder of a young healthcare startup still struggling to close our pre-seed, I have to work within my limits.” To circumvent the talent shortage, MyItura 's Chief Technology Officer, who is also an AI engineer, currently shoulders the bulk of the workload and collaborates with younger interns.
High cloud computing costs present another challenge for the business. ”Without our current subsidies, we’ll be spending over $5,000 monthly to train our own internal models and get our product to the user," Arogundade explained. To manage these costs, the company relies heavily on startup credits from organisations like Google, Microsoft, and MTN. This is a temporary fix. "The credits will finish in about a year," he added, noting that cloud costs will soon become a balance-sheet line.
While navigating cloud computing costs, fierce talent competition, and unstructured data are standard hurdles for companies building AI solutions on the continent, MyItura faces a different challenge in the low-income communities it serves. In these areas, the basic digital infrastructure required to host AI tools is often nonexistent. "A hospital does not even have a laptop or a desktop, so how do you deploy AI or EMR in the first place?" Arogundade asked. MyItura’s workaround is securing grant funding (from the Sonder Collective) to equip 35 field agents with tablets across Lagos and Ogun states. These agents help 230 primary healthcare centres (serving 20,000 patients) digitise operations without placing any financial burden on the hospitals. The company is also exploring USSD codes and Interactive Voice Response (IVR) systems to reach users in low-bandwidth areas.
Regulatory frameworks and collaboration are key to scale
Despite these hurdles, Arogundade expects AI adoption among medical practitioners to grow. Looking ahead, MyItura's next major step is completing the machine learning training for its Mediloan product, which will allow it to provide even faster credit responses using both internal and third-party data.
Achieving this product vision requires regulatory support. “The biggest thing for me is the talent subsidy," Arogundade stressed. "Making sure I can afford to bring people in when I need to." While current regulatory ambiguity allows for rapid innovation, Arogundade believes that establishing clearer frameworks around AI use, alongside enforced data interoperability standards across hospitals, would provide a much stronger foundation for sector-wide growth. He notes that more intentional, locally-led collaboration between the government, universities, researchers, and startups is essential to build a robust AI ecosystem across the continent.
–
If you are an African business adopting or developing AI technology, we want to hear from you . Fill out the survey and join 80+ other startups that have already shared their experiences as AI businesses in Africa. Your input will advise policymakers through the GIZ African Union AI Made in Africa programme.
