SECTION 4 · INNOVATION & GROWTH

Product

We build software for settings that range from busy urban drug shops to remote pharmacies with limited connectivity and thin margins. Every feature we ship has to be simple enough for a first-time user. And it needs to be robust enough to support everything from a single pharmacy sale to a nationwide health program.

In 2025, our product team set out to make every step of that cycle — including receiving stock, recording a sale, and getting reimbursed — easier and more accurate through automation. That meant reducing manual steps that introduce errors, adding safeguards that work quietly inside existing workflows, and laying technical groundwork for the AI-assisted tools we're developing for 2026.

2025 highlights

What we shipped this year

Serialization & Supply Chain Checks

Before medicines leave our warehouse, each unit is labeled with a unique code that follows it to the point of sale. For products sold through our health programs, our app automatically checks that code against third-party delivery records to confirm the product is actually in stock. Any discrepancies are captured automatically, and our system flags these sales for audit.

Stocking Shelves, Seeing the Full Picture

Private pharmacies source products from a wider range of suppliers than what you'd see in government supply chains. Manually entering each one is time-consuming and prone to small errors. This year we introduced two new features: First, we designed and deployed a tool that lets staff photograph a product's packaging and automatically extract key details (e.g., brand name, strength, active ingredient) using “optical character recognition,” a form of AI that reads text directly from images. Second, our app now autocompletes drug details as pharmacists type, drawing from our own product catalogue — now one of the most comprehensive records of what's sold across the private sector.

Faster Payments to Providers

As our patient programs expanded, our provider reimbursements moved through more hands, more spreadsheets, and a wider mix of mobile money platforms to reach providers in each country. In 2025, we redesigned Malipo, our internal payments system, to consolidate that work into a single portal. Once Madai audits and approves a claim, Malipo now batches and disburses reimbursements with far less manual handling. This reduces touchpoints, prevents reconciliation errors, and quickens turnaround for providers who depend on these payments to keep stock on the shelf.

2026 roadmap

What's next

Helping Pharmacies Forecast their Inventory Needs

For pharmacies running on thin margins, ordering the right products in the right amounts is critical. In 2025, we ran a hands-on pilot that showed we could accurately forecast inventory needs, with support from the Patrick J. McGovern Foundation. This work also gave us a clearer read on what would actually drive uptake: not just our mathematical model, but how that forecast reaches providers and how easy it makes ordering. In 2026, we're launching an improved way to plan inventory through the channels pharmacies already use (i.e., WhatsApp, the Maisha Meds app), with a PDF of order lists they can send straight to suppliers.

New Programs, New Patient Flows

Our app walks staff through the steps of each patient program, from screening to dispensing. In 2025, we redesigned priority workflows as our programs expanded — including customized, simplified flows for HIV prevention and malaria as part of Uganda's AMDIS program. In 2026, we will take this even further to support new initiatives such as HIV treatment pickups in Kenya and our first tuberculosis program in Nigeria.

AI Tools That Keep Humans at the Center of Care

Pharmacies and drug shops are often the first place a patient describes their symptoms to anyone, and whoever's behind the counter must make consequential decisions in minutes. We're building tools that use AI and automation to help them capture richer information, surface clinically relevant cues during a patient interaction, and strengthen the human judgment at the center of every visit.

Product Spotlight

AI

Maisha Meds built its reputation by bringing the latest tech into settings that have often been the last to receive it. AI is the next chapter of that work.

While foundational models have advanced so much in just a few years, bridging the implementation gap relies on everything else: the design choices, data pipelines, workflows, hardware constraints, connectivity assumptions, and user trust. Fortunately, the “everything else” is our specialty.

For years, our partnership with Audere has applied computer vision to verify malaria and HIV test results from within our Madai auditing platform. More recently, we've used character recognition to identify product packaging for stock entry, and we're now testing LLM-enabled tools to help providers elicit clinical details from their patients to support care decisions. Our aim is to evaluate the feasibility, adoption, effectiveness, safety, and impact of these tools on everyday provider workflows and patient care.

A number of principles guide this work, including:

01AI should support providers, not replace their judgment.
02We design user-driven tools with patient care as our north star.
03We experiment, learn, and scale thoughtfully the same way we do across our other programs.
04We contribute to the evidence base, especially where little evidence currently exists.
05Safety, patient privacy, and human oversight underpin every tool we develop.

This work is possible thanks to funders including the Patrick J. McGovern Foundation, and we look forward to joining The Agency Fund's 2026 cohort to take it further. Ultimately, Africa's community pharmacies and the people they serve should not be the last to benefit from cutting-edge innovation — and in fact, they can help shape it.