Artificial intelligence is slowly reshaping healthcare delivery in Nigeria, from pregnancy and immunization support to general healthcare. While one startup is connecting mothers with timely health information through a WhatsApp chatbot, another is providing general healthcare services with promises to overcome the overstretched healthcare infrastructure in a country with a limited workforce. Together, these tools reflect how Nigerian innovators are tailoring AI to the country’s most pressing health challenges.
“Hello and welcome to HelpHum chatbot service. This is your one-stop source for your pregnancy [MamaBot] and child immunization [VaxBot] information service,” that was the first message that popped up when HealthTabs tested the HelpMum WhatsApp chatbot. For millions of Nigerian women who often lack timely access to medical guidance, such messages could be the first step toward safer pregnancy and healthier children.
Nigeria continues to grapple with unacceptably high maternal and child mortality despite government interventions over the years. About 28.7% of all maternal deaths worldwide in 2023 were recorded in Nigeria, the figure representing 993 deaths per 100,000 live births. Overstretched healthcare infrastructure and limited manpower also deepen this crisis.
In 2024, 14.3 million infants did not receive an initial dose of the Diphtheria, Tetanus and Pertussis (DTP) vaccine; an additional 5.6 million infants were only partially vaccinated. Of the 19.9 million under-immunized children worldwide, about 55% live in just 10 countries, Nigeria inclusive, according to the World Health Organization (WHO).
The country remains one of the countries where maternal and neonatal tetanus remain a public health problem. This necessitates an urgent need for innovative solutions to strengthen maternal and child healthcare, vaccine delivery and uptake in Nigeria and other high-burden countries.
AI tools could help ease Nigeria’s health system pressure. With less than four doctors per 10,000 people, AI-powered assistants may bridge early diagnostic gaps, promote self-care, and support overburdened clinicians.
While government efforts show political will, they have yet to produce significant results. This is why technology-driven interventions are gaining attention. Though there have been reports of risks and failures of AI in healthcare, a quiet revolution is happening through the localized solution of a Nigerian health tech startup, HelpMum Africa. The AI-driven free chatbot offers an innovative way of bridging gaps in maternal and child healthcare. By automating reminders and guidance, the tool directly tackles vaccine defaulting, one of the major contributors to preventable childhood illnesses and mortality in Nigeria.
Mrs Aminat Kareem, who attended the antenatal clinic at Atelewo Primary Health Centre, Osogbo, said she did not know about the HelpMum chatbot but got health education on her pregnancy from the health workers at the PHC whenever she visited for antenatal clinic. She expressed her willingness to use an AI tool with HelpMum chatbot’s flexibility that could complement the education given by healthcare providers. “If the information were sent to me on WhatsApp, then I would be willing to try it. That would mean I can have access to information at any time, not only when I come here,” she stated.
The one-stop source for pregnancy and child immunization information
The chatbot is a two-in-one component that allows users to pick between pregnancy and child immunization services. Users can also switch to the other service at will and it provides information as prompted. During one of our trials of the app, we opted for VaxBot, and it displayed available services which included ‘vaccine information, location service, ask me your vaccination questions and add information about another child.’
HealthTabs picked vaccine information and an option to pick the chatbot brought a list of ages at which childhood vaccines are given, HealthTabs picked ‘14 weeks’ and the bot listed all the vaccines appropriate for the age, which includes Inactivated Poliovirus Vaccine (IPV), Oral Poliovirus Vaccine (OPV), Pneumococcal Conjugate Vaccine (PCV), Pentavalent Vaccine, and Rotavirus Vaccine.
“At 14 weeks, your child is due 5 ️ vaccines: IPV (dose 2 of 2); OPV (dose 4 of 4); PCV (dose 3 of 3); Penta (dose 3 or 3); Rota (dose 3 of 3). These are all routine childhood immunizations that the National Primary Health Care Development Agency strongly recommends (sic),” a message from the VaxBot read.
The VaxBot added that IPV protects your child against polio, a disease that can cause limb weakness and could eventually result in a total inability to move. It further added that there was no cure for polio, so vaccination was crucial to stay safe. The chatbot went on to add a little information about each vaccine, including their advantages. It further warned against non-compliance with the vaccines, their side effects and dosage were also listed. The message from the chatbot concluded with a promise to send a reminder when the next dose was due.
Comparison of the information provided by the chatbot with the National Programme on Immunization Schedule showed that the information provided by the chatbot was accurate.
When HealthTabs checked MamaBot, it asked if the user had registered for antenatal and it advised registration at a preferred medical provider when the response was negative.
It later asked for the age of the pregnancy as the chat progressed, HealthTabs replied ‘14’ and the chatbot immediately and correctly estimated the expected date of delivery. “This is second trimester and you might start feeling hungrier, more energetic, and less nauseous as early symptoms start to subside. At this point, your baby could be sprouting some hair on the body and eyebrows,” the information provided by MamaBot on a 14-week-old fetus reads in part.
On nutrition, it advises pregnant women to prioritize food items that give them calcium since it is necessary for the hardening of fetal bones that is occurring around this time. The chatbot also advised against strenuous exercises. “Do not perform jerky or twisting motions which can further strain your stressed-out joints. Do not do any exercises that affect your sense of balance or can cause trauma to your abdomen,” MamaBot advised.
A week later and without any prompting, MamaBot sent information about week 15 of the pregnancy. This time, it said the fetus was looking more like a baby as its eyes were moving to the front of the face. It added that pregnant women will start looking pregnant and gaining weight since their growing abdomen will become noticeable as their fetus becomes bigger at this time.
A pregnant woman, Sarah Azeez, who used the app following her introduction to it by HealthTabs, reported satisfactory service. “At 18 weeks it indicates that I may start to feel movements of the baby, the information is the same as what the nurses tell us whenever we go for antenatal care,” she stated while narrating her experience using the chatbot. First fetal movements, also known as ‘quickening’, usually occur around the 18th week of pregnancy in women who are pregnant for the first time; it occurs at about 16 weeks in women who have been pregnant before.
“If chatbot can help us reduce maternal mortality, it is welcome”
A Community Health Extension Worker, Idowu Lawal, while commenting on the chatbot, believed the chatbot would provide important health information to pregnant women and nursing mothers. “I checked the link you sent to me on WhatsApp and I say it worked well, it will really go a long way in helping our patients have information at their fingertip, we educate them each time they come for antenatal and child immunization, but knowing they can readily have the information at their disposal is a relief and welcome development,”
She, however, expressed concern that while health workers speak to the patient in their native language at the PHCs, there may be a language barrier to accessing the service by patients who aren’t literate. She added that those who do not have smartphones may not benefit from it too. “However, I will introduce it to my co-workers and our patients because I’m just getting to know that something like this exists. On immunization, some people refuse immunization for their children, not because of lack of information but just out of willfulness,” she concluded.
“Maternal mortality can be reduced to the barest minimum, though we primarily manage the patients we like to employ multidisciplinary approach so we are always ready to collaborate with anybody or anything that can help us achieve that goal. If a WhatsApp chatbot can help us reduce maternal mortality, it is welcome,” Dr Afolabi, a gynecologist and obstetrician, said while commenting on how the chatbot can be useful to them.
Dr Adeola Adebayo, a pediatrician, in her comment, confirmed the accuracy of the information provided by the application. “I cross-checked the information (on childhood immunization) from the chatbot, and I can confirm they are correct, so I think it’s safe for use,” she stated.
GabbiDoc: The AI Doctor Bringing General Healthcare to WhatsApp
When HealthTabs tested GabbiDoc, an AI-powered chatbot developed by Gabbi Health, the interface opened with a friendly greeting just like HelpMum’s. It went further with a promise to connect one with a real doctor in real time if need be. The chatbot aims to overcome the overstretched healthcare infrastructure and limited manpower challenges while ensuring improved patient outcomes.
“Hi there! I’m Kiki, your personal AI health assistant from GabbiDoc. I can help you understand your symptoms, get instant answers to health questions, and connect with a real doctor when you need one.”
GabbiDoc’s approach goes beyond targeted care like maternal or child health; it offers a wide range of services through WhatsApp. Users can consult the AI doctor, scan and review lab results, or chat with a licensed physician after paying a modest subscription fee of ₦1000 monthly, ₦2500 quarterly, or ₦10,000 annually. This low-cost, on-demand structure aims to make medical advice more affordable to Nigerians struggling with doctor shortages and hospital delays.
For instance, when HealthTabs described symptoms such as fever, jaundice, and upper abdominal pain, GabbiDoc quickly responded with a differential diagnosis list, which includes gallstones, cholecystitis, or hepatitis and was accompanied by simple explanations, home care tips, and guidance on possible tests, including Liver Function Tests, Abdominal Ultrasound, and Complete Blood Count. The chatbot also recommended consulting a gastroenterologist. In another session, when symptoms of chest pain were entered, it suggested likely causes such as costochondritis or muscle strain, advising rest and mild pain relief medications such as paracetamol or ibuprofen.
Beyond English, GabbiDoc can converse in Nigerian local languages, including Hausa, Igbo, Pidgin English and Yoruba, responding seamlessly when the user switches languages, an impressive localization feature that could enhance accessibility among non-English speakers. “Bẹẹni, mo gbọ́ Yorùbá. Bawo ni mo ṣe le ràn ọ́ lọ́wọ́?” (“Yes, I understand Yoruba. How can I help you?”) the chatbot replied during testing.
The chatbot also allows users to upload laboratory results for AI-assisted interpretation. When a sample test result was submitted, the chatbot provided a brief report, noting elevated potassium levels and a diagnosis of symptomatic bradycardia, along with safe explanations and recommendations to consult a doctor for confirmation.
While the tool provides instant support, its developers acknowledge that GabbiDoc is not a replacement for in-person medical care. Instead, it aims to serve as a first-contact assistant that guides users toward appropriate next steps.
Responses are based on validated information from WHO, CDC, and experts – Service providers
Taiwo Akintola, a staffer of HelpMum, while fielding questions on the inspiration behind the chabot was designed to help mothers experiencing some forms of constraint in accessing basic information on healthcare can get information from the comfort of their homes.
To ensure accuracy, Akintola said HelpMum works closely with healthcare experts who review all the information provided by the chatbot. “These professionals cut across a wide range of specialties such as obstetrics and gynecology, general medical practice, public health, amongst others. Also, we ensure accuracy by ensuring our data sources are accredited and certified by health organizations.” They added that they work with some of the best AI Engineers to ensure parameters such as precision, bias, and hallucination are catered for.
“Currently, HelpMum is National Data Protection Regulation (NDPR) certified and we uphold every measure of data security to ensure our user data is safe and secure. Also, we ensure the development of our innovations are on the most secure platforms regardless of cost. This is purely because we prioritize the data of our users. In addition, we follow key data security,” they stated while commenting on how the organization is ensuring data of users are protected.
They added that though the chatbot had already reached about 180,000 mothers, HelpMum sought to reach more women through partnership and adoption by state governments. “This goes beyond hitting targets and adding up to the current figures it would be an opportunity to scale and make impacts in the lives of mothers, infants and generally caregivers,” they concluded.
The Founder and Chief Executive Officer of Gabbi Health, Dr Kiki Diorgu, in an interview said GabbiDoc was born out of the severe shortage of medical practitioners and poverty that has caused a lot of people to neglect their health, with many ending up delaying treatments or turning to harmful alternatives. “Gabbi-Doc is a 24/7 AI Doctor on Whatsapp that was built to enable people access quick and adequate medical care, correct health information, and affordable health care and we chose WhatsApp because it’s already a part of our daily lives here in Nigeria. People don’t need to download anything new or learn how to use a fresh app,” he stated.
Dr Diorgu added that GabbiDoc’s responses are based on medical guidelines and data from reputable sources like the WHO and the Centre for Disease Control. “In addition, we have a review system where licensed healthcare professionals periodically vet the content to ensure it remains clinically sound and locally relevant. We also keep the system regularly updated to align with the latest medical standards and best practices,” he added while responding to questions on how they were ensuring accuracy in the output of the chatbot.
The Gabbi boss said the company safeguards patient data through end-to-end encryption, role-based access control, and multi-factor authentication. He added that its HIPAA-compliant cloud servers feature advanced firewalls and intrusion detection systems, while all activities on the platform are logged for accountability. The company also complies with NDPR standards and conducts regular security updates, vulnerability checks, and encrypted backups to prevent data loss.
He disclosed that they aim to expand access by offering the platform in major African languages and adding voice features for users with low literacy or limited internet access. He also added that the company is partnering with public health agencies and Non-Governmental Organizations to make GabbiDoc a frontline digital health educator and triage tool. The goal, he said, is to support rural communities with limited access to doctors, provide instant first-aid guidance, and make GabbiDoc a trusted, accessible first point of care across Africa.
“Responses Generated by AI are not Infallible” – AI Engineer
While HelpMum’s chatbot has been praised for accuracy and expanding access to maternal and child health information, experts warn that accuracy and reliability must remain a priority. Chisom Rutherford, an AI engineer, who spoke with HealthTabs explained that despite recent advances, AI systems are still prone to what he described as ‘hallucination,’ where they generate responses that may sound correct but are factually inaccurate.
“Current models may still inevitably generate responses that are inaccurate. The best safeguard is to ensure that the chatbot’s responses are drawn from trusted, verified data sources or fine-tuned with local health data that reflect Nigeria’s realities.” He added that users should understand the limitations of AI, noting that “responses generated by AI are not infallible, and mistakes could always happen.”
The engineer also weighed in on the glitches observed during user tests. He described these as engineering challenges rather than core AI flaws. “It’s an example of why healthcare apps need rigorous testing before being released to the public. These are issues that the right skills can resolve, and doing so will improve user trust and overall experience,” he explained.
On the sensitive issue of privacy, he stressed the need for stronger safeguards. He advised that personal identifiers like names, phone numbers, or addresses should not be fed directly into AI systems. Instead, such data should be anonymized before processing. He also recommended independent audits to ensure that health tech companies actually adhere to the strict privacy rules and policies required for processing healthcare data.
Rutherford added that he was optimistic about AI’s potential to ease Nigeria’s healthcare burden. With maternal and child mortality rates still among the highest globally, he believes tools like MamaBot, VaxBot and GabbiDoc can play an important role in bridging the gap created by weak infrastructure and shortage of health workers. “In the future, I envision AI agents reducing the burden on our already fragile healthcare system. Only critical cases will be escalated to a human, while AI handles less critical cases and offers health education at a fraction of the cost.” He however, cautioned that close human supervision is necessary for now to prevent harm from false information.
This report was produced with support from the Centre for Journalism Innovation and Development (CJID) and Luminate
So detailed
Thanks