What is the name of your organization?
eSHIFT Partner Network
What is the name of your solution?
FacilityPulse
Provide a one-line summary or tagline for your solution.
Open-source AI toolkit converting citizen Google reviews into actionable population health intelligence for health ministries in LMICs.
In what city, town, or region is your solution team headquartered?
Geneva, Switzerland
In what country is your solution team headquartered?
Switzerland
What type of organization is your solution team?
Hybrid of for-profit and nonprofit
Film your elevator pitch.
What specific problem are you solving?
Health ministries in low- and middle-income countries lack real-time intelligence to shift from reactive to anticipatory care. Existing surveillance systems are fragmented and slow — often taking weeks to surface crises that affected communities are already experiencing and documenting publicly. Meanwhile, citizens generate continuous health signals through platforms they use daily: Google reviews describing overcrowded clinics, drug stockouts, broken equipment, absent staff, and deteriorating hygiene. These signals are entirely unmonitored by health authorities.
Compounding this, the foundational data layer underpinning all health system decision-making — national Master Facility Lists — is chronically outdated, inconsistent, and incomplete. In Sri Lanka, official facility names frequently diverge from how those same facilities appear in Google Maps, reflecting naming drift and unauthorized third-party registrations. This cascades through every downstream system: outbreak surveillance, resource allocation, and coverage reporting.
This twin failure — missed community health signals and degraded facility data infrastructure — characterizes health systems across Africa and Asia. An estimated 4.5 billion people live in settings where health authorities cannot see community health experience clearly enough to act anticipatorily. The tools to close this gap already exist. They are simply not being applied to public health.
What is your solution?
FacilityPulse is an open-source toolkit that transforms publicly available citizen voice data - specifically Google Maps patient reviews - into structured population health intelligence for ministry decision-makers. It requires no new devices, no behavioural change from communities, and costs near-zero beyond API fees.
The toolkit runs on 3 components: N8N (open-source workflow automation), Odoo (facility registry and alerting), and a commercial AI LLM. Weekly, it pulls reviews for every registered health facility via the Google Places API, feeds them through an AI analysis layer that detects pattern - stockouts,staff absences, hygiene failures, overcrowding, sudden rating drops & stores a structured intelligence report against each facility. When the AI detects urgent concerns, it automatically generates prioritized alerts routed to relevant decision-maker: CMO, district health officer, or supply chain manager.
A second layer - MFL Constant Gardening - uses AI to continuously reconcile official facility registries against public data, detecting naming discrepancies, unauthorized registrations, and facilities present in public data but absent from official records.
Already running across 30 facilities in Sri Lanka, SOLVE funding will package this as a fully documented open-source toolkit — N8N workflows, Odoo free deploy, AI prompt libraries, and deployment guide - published under an open license.
Who does your solution serve, and in what ways will the solution impact their lives?
FacilityPulse directly serves two groups. The first is ministry of health leadership in LMICs - chief medical officers, district health officers, program managers, and supply chain administrators - who currently make consequential decisions based on delayed, incomplete official reporting. FacilityPulse delivers role-appropriate intelligence to their existing workflows without requiring new infrastructure or training.
The second and ultimately primary beneficiary is the populations these facilities serve. In Sri Lanka, the 30 pilot HIQI facilities collectively serve hundreds of thousands of patients annually. National scale across 1,700+ public facilities and hundreds of regulated private providers covers over 22 million people. Across LMICs broadly, an estimated 4.5 billion people depend on health systems that currently have no systematic mechanism for capturing community health experience.
These populations are underserved in a specific way: they generate health signals constantly through platforms they already use, but those signals never reach the officials responsible for their care. A patient posting a Google review about a drug stockout is contributing a public health data point that disappears into a commercial platform. FacilityPulse closes that loop — at scale, at low cost, and without surveillance risk, since all data sources are voluntarily and publicly contributed.
