How AI is helping Mozambique to boost vaccine decision-making
Twenty-five percent of children in Mozambique have never had a single vaccine. A new AI-driven data platform is helping to close these gaps.
- 4 November 2026
- 8 min read
- by Priya Joi
At a glance
- Nearly one in four children in Mozambique has never had a single routine vaccine. Key information is scattered across a dozen separate systems, making immunisation decisions challenging.
- An AI-driven data platform developed by Zenysis with Mozambique’s Ministry of Health, supported by Gavi, now brings it all together without replacing any existing system.
- In September 2026, the country announced it would expand district-level use of the platform from 15 districts to more than 80 across six provinces, building on its current use across all 11 provinces.
Being a vaccinator in Mozambique can mean trekking to remote, hard-to-reach communities and struggling to keep vaccines cold in places with no reliable electricity.
But one of the hardest problems is that of fragmented data sets. The data needed to plan immunisation services and allocate resources is scattered across systems that don’t talk to each other: one for vaccination records, another for vaccine stocks, a third for population estimates, each built for its own purpose and in its own format.
Now, that is starting to change.
In September 2026, Mozambique announced it would continue to expand the AI-driven Advanced Data Analytics Platform (PAAD), which the country brought in almost a decade ago, from 15 districts in Maputo City and Maputo Province to more than 80 districts across Maputo City, Maputo Province, Inhambane, Tete, Nampula and Zambezia, building on its use at the national and provincial levels.
The expansion is prioritising districts based on performance gaps and greater data-quality challenges.
The PAAD platform, co-developed by Zenysis Technologies and Mozambique’s Ministry of Health (MISAU), with support from Gavi, integrates and harmonises data from more than a dozen existing systems into a unified analytics environment, without replacing any of them.
It allows national and subnational teams to look across information that would otherwise remain fragmented or require laborious manual reconciliation and use it for vaccine planning, quantification, performance monitoring, data quality, logistics and resource allocation.
About 24% of children in Mozambique are zero-dose, meaning they have not received a single routine vaccine, according to the 2022–2023 Demographic and Health Survey. In a country facing a gap that large, a clearer view matters. It helps immunisation teams to see which districts are falling behind, and where to act.
One picture instead of a dozen
The PAAD platform has become a central part of how the country’s vaccine programme works day to day.
With more than a dozen critical health systems integrated into PAAD, national, provincial and district teams use it to track coverage, check data quality, monitor vaccine stocks and spot the areas where children are being missed.
Coverage of the third dose of the diphtheria, tetanus and pertussis-containing vaccine (DTP3), a key marker of how well routine immunisation is reaching children, increased from 55% of children in 2022 to 70% in 2023, according to UNICEF and WHO data. Closing the remaining gap means knowing exactly where it lies.
“PAAD gives teams a clearer view of vaccination activity, vaccine availability and other indicators in one place,“ says Gabriel Risk Martins, Chief Partnership Officer at Zenysis. “It reduces the need to compile data manually from separate systems, making it easier to review performance, investigate inconsistencies and prioritise follow-up.”
“PAAD has made my work much easier; I can now complete various analyses in a fraction of the time that would have taken over a day using Excel,” says Alberto Jacenau, a provincial immunisation officer in Zambezia province. “I am learning something new every day, and with the support of the Zenysis team, I am constantly overcoming previous knowledge gaps.”
The new expansion will combine district-level access with practical training and ongoing mentorship for provincial and district teams, focused on using the PAAD platform during routine performance reviews to identify data-quality issues, prioritise corrective actions and monitor follow-up.
It is backed by provincial champions, digital learning resources and data review processes: the things that keep a tool in use for the long term.
A bet made before the AI boom
Back in 2017, Gavi’s Innovation for Uptake, Scale and Equity in Immunisation (INFUSE) programme, which seeks out promising innovations and helps countries take them up, selected Zenysis as one if its Pacesetters for their innovative, open-source software, Harmony, designed to integrate data from fragmented health information systems.
Private-sector catalytic funding, channelled through INFUSE, paid for the first deployment of the technology in the country.
In 2019, tech multinational organisation Tencent provided US$ 1.7 million in funding. This early funding was important to allow the country to try a new approach without spending scarce domestic money to find out whether it worked.
“When INFUSE selected Zenysis nearly a decade ago, nobody was talking about AI for immunisation,” Bertrand Pedersen, Acting Head of Private Sector Partnerships & Innovation at Gavi, told VaccinesWork. “What we saw was the solution to a recurring problem: countries had a lot of data that couldn’t be used together.
“Catalytic funding from Tencent is what allowed Mozambique to test a new approach without putting scarce domestic resources at risk, and once the value was proven, the Ministry of Health took it forward on its own terms.”
This first deployment happened in 2019, when Cyclone Idai tore through central Mozambique, and cholera followed as floodwater contaminated drinking water. Data on the outbreak was scattered across multiple systems run by different partners, requiring manual reconciliation when every second mattered.
Zenysis worked with the National Institute of Health (INS) to bring information together into a virtual emergency control room. Rather than replacing the tools already used by government and partners, the platform connected and harmonised their data, resolving inconsistencies such as duplicated facility names and incompatible formats.
This gave response teams a more coherent and timely view of the outbreak, while allowing existing systems and workflows to remain in place.
The results were immediate. With the government using the platform to monitor the outbreak and implement targeted measures, cholera cases fell in Sofala Province from 400 a day to zero in under a month.
Based on this success, Zenysis partnered with the Ministry of Health to consolidate that emergency tool into a broader solution able to support routine programmatic health improvements, giving birth to PAAD. “The technology is agnostic,” says Risk Martins, “which is why it is able to serve the entirety of the health system, including outbreak response and immunisation.”
Have you read?
The ministry takes the lead
What happened next is the real test of any innovation. The Ministry of Health recognised the platform’s value and, with Gavi’s support, continued its partnership with Zenysis to expand the platform’s use in routine immunisation, where vaccination teams faced similar challenges in bringing data together for effective decision-making.
For the country's immunisation programme, PAAD’s combination of data integration and analytical tools means being able to see the whole picture.
By combining vaccination data with population projections and outreach data, teams can easily calculate coverage and dropout rates for any period, from national level down to a single health facility.
Built-in rules and alerts automatically flag districts or facilities where the numbers don’t add up, for instance where doses reported don’t match the vaccine available, so follow-up goes where it is needed.
Vaccine stock, open vials and doses given can be viewed together rather than in separate systems, helping to improve supply chain and funding usage efficiency PAAD also helps teams see where children are most likely to be missing out.
One analysis uses a practical proxy for zero-dose burden (children who received the BCG vaccine, but not the first dose of the combined DTP-HepB-Hib vaccine) to highlight, quarter by quarter, the districts where more children appear to be slipping through the gaps.
Since the outset, PAAD has been co-developed by the Ministry of Health and Zenysis, supporting a gradual increase in national responsibility for the platform. The contract supporting the expansion announced in September 2026 builds on this process, equipping EPI staff to train and support colleagues as district-level use expands.
The longer-term aim is to extend PAAD’s use across the health system and progressively strengthen national ownership. “The important change PAAD introduces is not that Mozambique lacked these data before,” Risk Martins explains. “The data already existed, but in different systems and formats. PAAD brings them into a shared environment where they can be analysed together.”
“Since I was introduced to PAAD through Zenysis in 2023, my perspective on data analysis and utilisation has shifted significantly,” says Admiro Fernando Martins, Monitoring and Evaluation at the Ministry of Health. “Today, PAAD is my primary work platform; I use it daily to streamline analyses, integrate data from various systems, and monitor multiple variables from a centralised location.
“More than just a technological tool, it has become a vital ally in monitoring and evaluation, facilitating the interpretation of information and supporting evidence-based decision-making.”
What comes next
For Gavi, Mozambique shows how innovation partnerships can work: the private sector brings the innovation, catalytic funding lets the country test it and the government then takes ownership and carries it to national scale.
“The real measure of success isn’t the pilot,” says Pedersen. “It’s the moment a government decides to own and scale an innovation itself. That’s what happened in Mozambique, and it’s the model we want to repeat: the private sector brings innovation, catalytic funding lowers the risk, and countries take it to scale.”
The platform is also laying the groundwork for the future. “Because PAAD already integrates and harmonises data from multiple sources, it provides the underlying data foundation needed to power more advanced capabilities such as forecasting, anomaly detection, and various types of predictive analyses in the near future,” says Risk Martins.
