Simprints Uses Fingerprints and AI to Prove Vaccines Reach the Right People

A health programme can count how many vaccine doses it delivered and still not know which people actually received them. Paper immunisation cards get lost, names get misspelled, and in communities where many people share similar names, children can be counted twice or missed entirely. For the roughly 1 billion people worldwide who have no formal proof of identity, this gap decides whether they get their second vaccine dose, their antenatal check, or their HIV medication refill. A nonprofit called Simprints closes that gap with fingerprint and facial recognition technology running on ordinary phones, and in 2024 it verified 551,000 health services, from vaccine doses to surgeries, proving exactly who received each one.

Born at Cambridge, Built for the Field

Simprints began as a research project at the University of Cambridge in the United Kingdom. Its founders asked a simple question: what if a health worker in a remote village could identify a patient as reliably as a border agent at an airport? The answer they built is a rugged fingerprint scanner and matching system designed for conditions that consumer biometrics were never made for: dust, heat, intermittent electricity, hands roughened by farm work, and no reliable internet connection.

The system works offline on low cost Android phones. A health worker registers a person once by scanning their fingerprints or, for newborns too young for reliable fingerprints, capturing a facial image. From then on, every visit is verified with a quick scan. The match happens on the device itself, which means no connectivity is required and no data needs to travel to a distant server. Simprints describes its biometric ID platform as the world's first built specifically for global development projects, and it releases its technology as open source so partners can inspect and audit how it works.

The Vaccination Verification Problem

Immunisation is one of the most cost effective health interventions ever devised, but its success depends on follow up. Most vaccines require multiple doses at specific intervals. When a caregiver brings a child to a clinic and the paper card is missing, the health worker must guess: has this child already had their first measles dose, or not? Guessing wrong in either direction has costs. A missed dose leaves a child unprotected. A duplicate dose wastes scarce vaccine supply.

Simprints worked with Gavi, the vaccine alliance, on exactly this problem. The project linked children to their vaccination records through biometrics so that immunisation history could be recovered with a fingerprint scan even when every paper record was gone. In later work, Simprints developed camera based, contactless facial recognition so that a single tablet photo of an infant could anchor a digital health card for the first years of life, with no paper to be filed or lost. For newborns, who receive several vaccinations in their first months, this continuity matters enormously.

From Pilots to Half a Million Verified Services

The numbers now speak for themselves. In 2024 Simprints delivered 551,000 biometrically verified health services across its operations, including vaccines, surgeries, and deworming treatments. The year before, reporting across eight major projects in Bangladesh, Ethiopia, Ghana, and Uganda, the organisation reached a total of 382,943 people with services including vaccines, medicines, and aid. The growth between those years reflects a shift that is rare in global health technology: moving beyond pilots to routine, at scale use embedded in government and NGO programmes.

The range of applications keeps widening. In Ethiopia and Bangladesh, biometric verification supports maternal health and nutrition programmes. In Ghana and Uganda, it underpins immunisation and primary care registries. In humanitarian cash transfer programmes, the same technology confirms that aid reached the intended recipient rather than someone collecting on their behalf. What links these use cases is the core promise: a single trusted record of who received which service, when, and where.

Privacy by Design

Biometrics in vulnerable communities raises obvious risks. A database of fingerprints belonging to refugees or people living with HIV could cause serious harm if it leaked or were misused. Simprints addresses this with a set of design commitments: biometric data is encrypted, matching can happen locally on the device, partners own and control their data, and the platform is built so that a stolen phone does not become a stolen identity registry. The open source approach means independent auditors can verify these claims rather than take them on faith.

The organisation also argues that biometrics done carefully can protect privacy compared with the alternative. A stack of paper registers listing names, diagnoses, and addresses passes through many hands and sits unlocked in clinics. An encrypted biometric record with strict access controls can be far harder to exploit. The goal, as Simprints frames it, is not surveillance but inclusion: giving people without documents a way to prove who they are so the system serves them.

Why Verification Matters More Than Counting

Global health programmes are awash in estimates. Coverage rates are modelled from household surveys, administrative counts are adjusted for error, and funding decisions rest on numbers with wide uncertainty. Biometric verification replaces some of that guesswork with ground truth. When Simprints says 551,000 services were verified in 2024, it means 551,000 instances where the identity of the recipient was confirmed against their biometric record at the moment of service.

That kind of certainty changes how programmes learn. Duplicate registrations surface immediately. Dropouts, patients who disappear between doses or visits, become visible as specific people whom a health worker can follow up with. Supply planning improves because demand is measured rather than estimated. In a world where a billion people are effectively invisible to the systems meant to serve them, the humble fingerprint, matched by an algorithm on a $100 phone, is quietly making them count.