Frequently asked questions

The questions we get most

From mothers, community health providers, clinicians, donors, NGO partners, and government stakeholders.

The platform

How Maternanet works

Maternanet helps the health system identify pregnant women who are at high risk of complications weeks before those complications occur. We do this by giving community health providers digital tools to screen mothers at home, running that data through a predictive risk model, and sending an automatic alert when a mother's risk profile crosses a threshold. The CHP then makes a priority visit or coordinates a timely clinic appointment before an emergency develops.

Think of it as a smoke detector for maternal health — except instead of detecting a fire already burning, it detects the conditions that make a fire likely. We also help remove the financial barrier to acting on that alert through our Care Now, Lipa Later microfinancing partnership.

We are not a telemedicine app. We are not a symptom checker. We are the predictive intelligence layer that operates upstream of every other intervention.

Our Composite Women's Health Index™ synthesizes 85 variables across three categories:

  • 47 clinical variables — obstetric history, blood pressure, haemoglobin, gestational age, prior complications
  • 23 socio-economic variables — household income, insurance status, distance to facility, prior care-seeking behaviour
  • 15 environmental variables — county health infrastructure, seasonal road access, local clinic capacity

Validated against clinical outcomes and WHO checklist-based scoring. Recall (sensitivity): 79% — 21pp better than standard care. Precision: 81% — providers are not overwhelmed by false alerts.

Critically, the model explains its reasoning. A provider sees which specific indicators are driving the score, enabling clinical confidence rather than blind algorithmic compliance.

This is by design, not an afterthought. 60% of our users access via SMS or USSD on feature phones with no data connection.

  • Edge computing: Compressed ML algorithms run locally on rugged tablets and clinic PCs — risk scoring without cloud access
  • Data caching & Syncing: Screenings logged locally with timestamps, syncing bidirectionally when any connection — even 2G — becomes available
  • SMS/USSD interface: Mothers receive risk alerts, reminders, and health nudges via basic text — no smartphone or data plan required
  • 79% recall — validated in field conditions below 1kb/second connectivity

Patent filed: "Distributed Healthcare ML with Progressive Sync" — Kenya Patent Office, Application #2024/789.

No. Maternanet operates on a strict human-in-the-loop principle. Our model only flags and recommends — it never decides or acts autonomously. Every alert generates a human action: a CHP visit, a clinical escalation, a facility appointment. The clinician always makes the final call.

In a context where Africa faces a shortage of 6.1 million healthcare workers, our goal is to make every CHP significantly more effective — helping them prioritise their caseloads intelligently and focus their limited time on the mothers who need them most.

Our model explains why it flagged a mother — not just that it did. Trust is earned through transparency, not assumed through authority.

Maternanet operates on a Two-Channel Architecture to integrate natively with Kenya's national DHIS2 system and EMRs while securing direct depth data:

  • Channel A (OpenHIM): Clinical data routes directly to government eCHIS networks without manual entry.
  • Channel B (Maternanet Direct): Proprietary socioeconomic & depth data is gathered via direct maternal consent (QR/USSD).
  • County health departments access population-level dashboards. Generates automatic SHA claims reports.

Setup: 2–4 weeks including staff training. One-time $399 fee + $85/month facility licence. Contact info@maternanet.com for a free trial.

Impact & evidence

What the data shows

Across 4 active Kenyan counties (Nairobi, Narok, Machakos, Kajiado) over 18 months of deployment:

+21ppANC completion
−15%Missed appointments
77%6-month retention
12+Lives saved/yr
$8,100Cost per life saved

Validation: ANC data cross-referenced against county DHIS2 records and independent clinic registers. Our quasi-experimental evaluation uses difference-in-differences methodology with matched comparison sites and continuous model improvement. A peer-reviewed paper is in preparation with a Kenyan university partner for Year 3–5 validation.

Currently $8,100 per life saved, projected to fall to $650–$1,300 by 2028 as scale increases. The WHO estimates skilled birth attendance programmes cost $4,000–$12,000 per maternal death prevented — our model is competitive now and improves dramatically with volume.

Early intervention also reduces average care costs by up to 70% by treating complications early rather than managing emergencies.

Every $1 invested in Maternanet generates $7–$20 in long-term economic benefits — reduced emergency care costs, maternal outcomes, and household poverty prevention.

Yes — and this is the clinical foundation of Maternanet. Machine learning risk models for pre-eclampsia and obstetric complications have been validated in peer-reviewed literature (BioMed Central, WHO Digital Guidelines) showing clinically useful discrimination using routine maternal data. Risk profile simulation for prenatal risk monitoring is an emerging but evidence-supported methodology.

Our model identifies the combination of signals that, in aggregate, indicate elevated risk weeks before clinical symptoms manifest — giving CHPs and clinicians a window for intervention that standard screening misses entirely.

Community & CHPs

The Digital Doula network

A Digital Doula is an existing community health provider (CHP) equipped by Maternanet to conduct community-level screenings, follow up on flagged high-risk mothers, and bridge the household-to-clinic gap where most maternal deaths happen.

  • 10-day intensive bootcamp: clinical basics, app usage, device handling, escalation protocols
  • Continuous digital refreshers accessible offline on tablet or SMS
  • Quarterly in-person workshops for peer learning and protocol updates
  • Maximum 45-60 mothers per CHP with performance-based incentives

Currently 410 active Digital Doulas across 4 counties, with a 96% retention rate.

Most digital health interventions fail here. We built Maternanet from within the communities we serve — our founding team comes from these regions and understands that trust is earned through presence, not technology.

  • Co-design: Features shaped in local workshops before they are built, not after.
  • Digital Doula ambassadors moderate peer forums — the platform is an extension of trusted social networks.
  • Local language delivery: SMS nudges in Swahili and local dialects.
  • Federated learning: models trained on-site without raw personal data leaving the community.
  • Explainable Recommendations: Providers and mothers see why the model flagged a risk, not just that it did.

Our 77% weekly engagement and retention rate — in communities with historically low digital health adoption — is the evidence this approach works.

Detecting risk without removing the barrier to acting on it is not a solution; fortunately we realized this earlier on. Maternanet addresses the financial barrier through two mechanisms:

  • Care Now, Lipa Later: Microfinance institution partnerships providing clinically-authorized emergency credit for predictively identified high-risk interventions — accessible at the point of the alert, but requiring mandatory Clinical Officer confirmation to prevent debt cycles.
  • Micro-saving wallets: Mothers save incrementally via mobile wallet — as little as KSh 30 per deposit — earmarked for delivery costs, reducing reliance on emergency borrowing

Our B2B revenue model subsidises care at the community level. The mother pays nothing extra. We take from the institutional top of the pyramid and give to the bottom.

Partner & fund

Working with Maternanet

Integration follows a four-step pathway:

  • Assessment (1–2 weeks): IT readiness, EMR/DHIS2 setup, catchment area characteristics
  • Integration and setup (2–4 weeks): Technical integration, tablet/PC installation, staff training
  • Digital Doula deployment: Existing CHPs from your community are equipped with our predictive tools
  • Go-live: Real-time dashboard showing population risk distribution, referral rates, outcome tracking

Pricing: $85/month facility licence + $399 one-time setup. Revenue-sharing partnerships available. Contact info@maternanet.com.

Three primary partnership models:

  • Programme co-deployment: Embed Maternanet within your existing maternal health programmes as the predictive intelligence layer
  • Population health analytics: License our data dashboards for evidence generation and policy advocacy — from $25,000/project
  • Academic research: Partner on impact studies including our upcoming peer review literature — academic licensing from $15,000/study

Early-mover partners in Uganda, Tanzania, Rwanda, Nigeria, and Ghana receive preferential data access and co-authorship opportunities on evidence outputs. Contact info@maternanet.com.

Currently deploying a $550,000 strategic capital round (through Q1 2027) to build the data foundation that unlocks high-margin DaaS revenue streams at scale by 2028–2029:

  • Technology enhancement — model refinement, autonomous agent pilots (36.8%)
  • First-mile logistics — CHP training and rural deployment (26.3%)
  • Partnership integration — county and facility expansion (15.8%)
  • Equipment and hardware — rugged tablets, diagnostic kits (10.5%)
  • Working capital (10.5%)

Trajectory: $95,897 actual 2025 revenue → $536,303 projected 2029. EBITDA break-even Q4 2026. ROI visible 18–25 months post-deployment. Contact owinoaketch@maternanet.com for the full investor pack.

No. Three diversified, reinforcing revenue streams:

  • B2B facility licensing — current primary revenue
  • Data-as-a-Service — activating 2026, government and NGO analytics contracts
  • Consumer and financial services — activating 2027, premium subscriptions and Care Now, Lipa Later revenue share

Grants and impact investment accelerate deployment — they do not sustain it. 20%+ net margins projected by 2028. Commercial success and social impact grow in the same direction: more enrolled mothers → better model → lower cost per outcome.

Data & privacy

How we handle sensitive health data

  • End-to-end encryption across all data pipelines with local data sovereignty — data collected in Kenya stays in Kenya
  • Federated learning: models trained on-site at local clinics without raw personal data leaving the facility
  • Kenya Data Protection Act (2019) compliance: Full alignment including data minimisation, purpose limitation, and subject access rights
  • GDPR-equivalent standards applied for all international data partnerships
  • Community consent frameworks: Every enrolled mother and CHP consents explicitly, with clear explanations in local languages

ODPC registration and DPIA approval in progress. KMPDC telehealth certification expected Q3 2026.

Whether our predictive tools qualify as Software as a Medical Device (SaMD) under Kenya PPB and WHO frameworks is a question we are actively working through — proactively, not reactively.

Current regulatory status:

  • NACOSTI Research Permit — already obtained
  • KMPDC digital health certification — in progress, expected Q3 2026
  • ODPC Data Controller registration — in progress
  • DHA Digital Health License — application underway
  • Kenya Health Data Governance Framework (2024) — compliance in progress
Scale & markets

Where Maternanet is going

  • Phase 1: 0 to 1 (2025–2028) — Kenya data foundation: 8+ counties, 80+ facilities, 500+ CHPs. Build national policy evidence base.
  • Phase 2: 1 to 100 (2028–2030) — National scale: Hub & Spoke expansion across Kenya, full DHIS2 and EHR integration, county cost-sharing.
  • Phase 3: 100 to Scale (2031–2035) — Pan-African entry: Uganda, Tanzania, Rwanda, Nigeria, Ghana.

Long-term vision: 1M+ women across 5 countries, Kenya achieving fewer than 150 maternal deaths per 100,000 live births by 2030.

Our model is transferable wherever three conditions exist: a community health worker structure, 2G mobile coverage (minimum), and a maternal mortality challenge driven by detection delays and access barriers. That describes most of sub-Saharan Africa.

Deployment requirements:

  • A local implementation partner with existing community health networks
  • A Ministry or county health department willing to sandbox with us
  • 6–12 months for predictive model localisation, consent framework adaptation, and regulatory clearance
  • Minimum cohort of 50 CHPs and 2 facility partners to build data density for model accuracy

We are actively seeking early-mover partners in Uganda, Tanzania, Rwanda, Nigeria, and Ghana. Contact info@maternanet.com with your context.

Maternanet is built to demonstrate SDG 3 achievement through a replicable, cost-effective model — not just aspire to it.

Primary alignment: SDG 3 (Good Health and Well-being), SDG 5 (Gender Equality), SDG 10 (Reduced Inequalities)

Secondary alignment: SDG 1 (No Poverty — reduced catastrophic healthcare costs), SDG 8 (Decent Work — CHP jobs), SDG 9 (Innovation and Infrastructure), SDG 17 (Partnerships)

We are not making incremental improvements to a broken system. We are demonstrating that technology-enabled, community-driven care can achieve SDG 3 targets cost-effectively — and creating a replicable model for the global South.
i The content in this page is structured as an FAQ page and may include content that has since been updated.

Still have questions? Let's talk.

We respond to every message within 48 hours — and yes, a human reads every email.