Skip to content
Cleampact
A man has his blood pressure taken at an outdoor community screening.

AI-Powered and Solar-Driven Disease Prediction and Prevention Detection Device for Local Communities.

LINDA gives Kenya's Community Health Promoters early warning of climate driven disease 2 to 6 weeks early warning before an outbreak and the prevention tools to protect the most vulnerable before the first case appears

How It WorksPartner With Us
LINDA sensor
Temperature
28.4 °C
Humidity
78 %
Air Quality
41 PM2.5
Water Safety
12 NTU

  • 107,000

    CHPs LINDA alerts across Kenya

  • 75%

    Reduction in severe malaria with timely intervention

  • 2–6 wks

    Early warning before outbreak onset

  • KES 350

    Monthly subscription per device

  • 3,091

    Household visits by CHPs during the founding intervention

Mission
To equip community health workers with early warning and prevention tools for climate-driven disease — protecting the most vulnerable, reducing disease burden, and safeguarding household income in rural communities.
Vision
A Kenya where community health workers receive early warning of every climate-driven outbreak and have the prevention tools to protect every vulnerable family before disease arrives.

Kenya has the data. The gap is the connection.

A mother holds her baby steady while two community health workers treat the child at a rural health post.
The households the data never reaches in time

Weather services collect climate data. DHIS2 tracks disease. 107,000 community health workers visit homes across 47 counties. None of these are connected.

Weather data stays in meteorological offices. Disease data stays in health facilities. Community health workers visit on a schedule, not on a risk signal. By the time a family knows there is an outbreak, a child is already sick, a pregnant woman is already at risk, an elderly parent is already deteriorating.

LINDA is the first system in Kenya to close this gap — combining climate sensors, AI disease prediction, and community health worker alerts into one real-time action loop.

Published evidence

  • 288%

    Surge in malaria cases in Nandi County following unusual rainfall and temperature shifts.

    Otieno et al., Malaria Journal, 2024

  • 12,666

    Cholera cases and 179 deaths in Kenya in 2022–2023, driven by flooding and water contamination.

    Onditi et al., PLOS Global Public Health, 2026

  • 27,000

    Deaths annually in Kenya from household air pollution caused by cooking fires.

    KEMRI, 2026

Who the alerts prioritise

Children under five
Highest mortality from malaria, pneumonia, and diarrhoeal disease.
Pregnant women
Malaria causes miscarriage, low birth weight, and maternal death. Heat stress triggers premature labour.
The elderly
Reduced ability to regulate body temperature. Compromised lung function worsened by cooking fire smoke.
Households cooking indoors on firewood
Chronic smoke exposure raises pneumonia risk for every family member.

From climate signal to community action in four steps

No new infrastructure. LINDA works on any phone, on 2G signal, in any community Safaricom reaches.

Community health promoters in blue vests gathered around a water container, demonstrating safe water handling to a group of women.
Step 04, in South Imenti

  1. 01

    Sensor reads the climate

    LINDA's solar-powered device reads temperature, humidity, air quality, and water safety every hour from a school or health centre in the community.

  2. 02

    AI predicts the outbreak

    A model trained on 30 years of East Africa climate data and 15 years of Kenya disease records detects patterns that have historically preceded outbreaks of malaria, cholera, heat stress, and respiratory disease.

  3. 03

    CHP receives a Swahili alert

    The nearest community health worker receives a Swahili SMS or WhatsApp message naming the disease risk, the climate trigger, affordable prevention steps, and which households to prioritise — 2 to 6 weeks before an outbreak starts.

  4. 04

    Community is protected

    CHPs visit households, distribute bed nets and prevention materials, brief school health clubs, and refer high-risk individuals early. The WhatsApp chatbot lets CHPs report back and ask questions in real time.

The warning windowA climate signal — rainfall, temperature and humidity — rises and peaks several weeks before the resulting case load peaks. LINDA alerts health workers during the gap between the two, two to six weeks before the outbreak starts.2–6 weeks to actClimate signalCasesSensor detectsFirst cases present
Schematic of the mechanism, not plotted data. Lead times vary by disease and community — see the warning times on each disease below.

Four climate-driven threats. One early warning system.

Warning time

Same day6 weeks

  • Malaria

    3–6 weeks early warning

    Triggered by rainfall, temperature, and humidity.

    LINDA predicts malaria transmission risk 3 to 6 weeks before peak season. Alerts prompt CHPs to distribute bed nets and Seasonal Malaria Chemoprevention medicines to children under five, pregnant women whose pregnancy may be complicated by malaria, and any household with no prior immunity.

  • Heat stress

    Same-day alert

    Triggered by temperature above 35 °C with humidity above 60%.

    Affects children under five, the elderly, and pregnant women. LINDA sends same-day alerts when dangerous heat–humidity combinations are detected.

  • Respiratory disease

    3–7 days early warning

    Triggered by cooking firewood and charcoal smoke, in cold and rainy seasons.

    The PM2.5 air quality sensor detects dangerous particulate levels 3 to 7 days before respiratory illness peaks. CHPs reach households with newborns, elderly, and pregnant women.

  • Cholera and diarrhoeal disease

    Early contamination detection

    Triggered by flooding, water contamination, and temperature.

    The water quality sensor detects turbidity and contamination at community water points after flooding. CHPs distribute water purification kits before cases appear.

LINDA Alert

WhatsApp · Cleampact Health System

Malaria Risk — High

Habari Purity. Hali ya hewa inaonyesha hatari ya malaria katika wiki 3 zijazo katika eneo lako.

Joto
28 °C
Unyevu
78 %
Mvua
Nyingi
  • Tembelea kaya zenye watoto chini ya miaka 5
  • Gawanya chandarua cha mbu
  • Toa dawa za kuzuia malaria kwa watoto
  • Rejesha wagonjwa wenye homa haraka

2G

The whole system fits in one message.

Everything a community health worker needs — the risk, the trigger, the households to visit — in the language they work in, on the phone they already carry.

Swahili first
Alerts are written and tested in Swahili by CHPs from South Imenti. Every recommendation is locally actionable and affordable.
2G SMS and WhatsApp
Works on any mobile phone. No data plan required for SMS alerts.
Two-way chatbot
CHPs reply to alerts, report field findings, and ask clinical questions. Every interaction improves the AI model's accuracy for the next season.

LINDA was designed from a lived experience.

“During a diarrhoeal outbreak at Nkubu, five in every ten children in our ward had the same illness. We trained 25 community health workers, 18 of them women, from the local community on prevention, referral criteria, and danger signs. They made 3,091 household visits. They broadcast in the local language on local radio. Within two months, admissions fell to one in ten. Sick children presented early, without complications.”

Dr. Dorris Mwangi

Co-founder and Clinical Lead, Cleampact

Community health workers seated in a circle on the grass outside a building during a training session.
Training the community health workers who did the visiting
25
CHPs trained18 of them women
3,091
Household visitsOne outbreak, one ward
5 in 10
Children ill at the startBefore the intervention
1 in 10
Children ill afterWithin two months

Prevention does not end with fewer cases.

When LINDA's early warning prevents one malaria hospitalisation, the impact ripples far beyond the clinic.

Household income protected
A single malaria hospitalisation costs KES 5,000–10,000 in treatment, transport, and lost income. An avoided hospitalisation keeps that money in the family.
Time returned to families
A mother who does not spend four days in hospital keeps four days of productive time — farming, trading, caring for other children.
Children stay in school
A child who does not get sick stays in class. School absenteeism during malaria season is tracked by the pilot as an independent impact measure.
Women earn more
75% of Kenya's CHPs are women. LINDA's earnings model — KES 200 per verified action, KES 500 per referral — directly increases income for women health workers.
Stronger health systems
Fewer patients reaching the HDU. Health workers can focus on acute care. Facilities are less overwhelmed during peak disease seasons.
Climate resilience built over time
Each pilot cycle retrains the AI model. Each new rainy season adds to the dataset. Communities build accumulated climate-health intelligence that grows more precise every season.

LINDA is built for partnership.

Three ways to work with LINDA. Each one starts with a conversation about the county, the season, and the households you already reach.

  • County governments

    Real-time community health intelligence for your county.

    LINDA gives county health management teams a live dashboard showing climate risk levels, CHP alert activity, and disease presentation trends — all linked to existing DHIS2 data.

    KES 350 per device per month

    Talk to us

  • Funders and investors

    Verified, real-time community health impact data.

    Disease risk trends, CHP intervention rates, and outbreak prevention outcomes — for reporting, research, and programme design.

    Pilot plan available on request

    Request the pilot plan

  • Hospital networks and NGOs

    Extend your community health reach with early warning.

    LINDA works alongside existing CHP networks and facility-based care. Alerts tell CHPs exactly when and where to act, before patients arrive at your door.

    Works with your existing CHP network

    Explore partnership

Partners and supporters

  • Mastercard Foundation
  • Zurich Foundation
  • Resolution Project
  • One Young World

Built by the clinicians and health workers who lived the problem.

  • Dr. Dorris Mwangi

    Dr. Dorris Mwangi

    Co-founder and Clinical Lead

  • Shannon Simiyu

    Shannon Simiyu

    Co-founder and Software Developer

  • Dr. Betha Lorah

    Dr. Betha Lorah

    Field Operations

  • Dr. Glory Mbone

    Dr. Glory Mbone

    Public Health Data Analyst

And the field team

Community Health Promoters, South Imenti

Community health promoters taking a man's blood pressure at an outdoor screening.

LINDA’s Kiswahili alerts are written, tested, and corrected by the CHPs who deliver them. Every message has to work at a doorstep before it ships.

25
CHPs trained in the founding intervention
18
of them women, from the local community
3,091
household visits they made
dorysmwangi@gmail.com

Six Sustainable Development Goals LINDA contributes to.

  • Sustainable Development Goal 1: No Poverty

    Household income protected when a hospitalisation is avoided.

  • Sustainable Development Goal 3: Good Health and Well-Being

    Climate-driven disease early warning and prevention.

  • Sustainable Development Goal 5: Gender Equality

    Income and standing for the women who make up 75% of Kenya's CHPs.

  • Sustainable Development Goal 10: Reduced Inequalities

    Health intelligence equity for marginalised communities.

  • Sustainable Development Goal 13: Climate Action

    Community-level climate resilience and health adaptation.

  • Sustainable Development Goal 17: Partnerships for the Goals

    County governments, funders, hospitals, and NGOs acting on one signal.

Questions we are asked most.

Not answered here? Write to us and a member of the team will reply.

dorysmwangi@gmail.com

  • LINDA predicts malaria (3–6 weeks early warning), heat stress (same-day alerts), respiratory disease from poor air quality (3–7 days warning), and cholera and diarrhoeal disease triggered by water contamination and flooding. Each prediction is based on climate conditions that have historically preceded outbreaks in that specific community.

  • Community Health Promoters — Kenya's government-deployed community health workers. LINDA works with the 107,000 CHPs already deployed, already paid, and already trusted by the communities they serve. No new workforce is needed. CHPs receive alerts via Swahili SMS or WhatsApp on any basic phone, on 2G signal.

  • Climate-driven disease does not choose who gets sick. LINDA's alerts prioritise households with children under five, pregnant women, elderly members, and families cooking indoors on firewood or charcoal. But the early warning benefits everyone in the community.

  • The founding evidence is real. During a diarrhoeal outbreak at Nkubu in Meru County, 25 local CHPs conducted 3,091 household visits and broadcast in the local language on local radio. Admissions fell from 5 in 10 children to 1 in 10 within two months. No child needed the HDU. LINDA automates and scales exactly that model.

  • Email us at info@cleampact.org. We welcome partnerships with county governments, hospital networks, NGOs, research institutions, and impact investors.

Ready to protect your community?

Tell us which county you work in and we will send the pilot plan.

Partner With UsFollow The Pilot