{"resourceId":"rio-connect-rio-ai-whatsapp-postnatal-outreach-data-lake-20261110","versions":[{"version":"external-493c06d7d143d5691f2845be8a3df377cf57d3c971daeb17dbd685b7f81a358e","resource":{"id":"rio-connect-rio-ai-whatsapp-postnatal-outreach-data-lake-20261110","title":"Rio de Janeiro uses AI and WhatsApp outreach on an integrated city data lake to find residents missing public services, starting with postnatal care","organization":"Rio de Janeiro City Hall (Connect Rio; Municipal Health Department; Iplan Rio)","sector":"Municipal public health and social services","geography":"Rio de Janeiro, Brazil","publishedAt":"2026-11-10","publicationDate":"2026-11-10","eventDate":null,"sourceName":"Journal of Medical Internet Research (News and Perspectives)","sourceLabel":"Independent journal news reporting","sourceUrl":"https://www.jmir.org/2026/1/e115514","evidenceClass":"independent-reporting","outcomeClass":"emerging","topics":["data-security","operating-model","accessibility-workforce"],"finding":"Since May, Connect Rio has used AI on a city data lake linking health, social services and education records to identify residents missing services and contact them via WhatsApp. In a two-month postnatal pilot, 300 discharged mothers were checked daily for warning signs; 15 were referred back to maternity services. The program is expanding to 1,500 women and, by December 2026, 11 maternity hospitals and 242 clinics. City hall credits it with higher welfare uptake and a roughly 10% attendance gain in the poorest neighborhoods.","sledRelevance":"U.S. county health and human services agencies weighing integrated data hubs and proactive outreach (postpartum Medicaid, benefits uptake, school absence) can see a working municipal model, though transferability is limited by Brazilian law, WhatsApp ubiquity and a single city-run system.","evidence":"JMIR correspondent reporting (Nov. 10, 2026) quoting Rio undersecretary for health Renato Cony, Iplan Rio president Thiago Trabach and former CTO Joao Carabetta, with pilot counts and expansion targets.","architectureImplications":"Centralized cloud data lake linking roughly 100 million consultations across services, a WhatsApp outreach tool, and nurse dashboards that show only action-relevant data per case.","governanceImplications":"Role-limited data access per team; outreach triggers human nurse follow-up rather than automated clinical action; the stated 30-35% maternal mortality reduction is an aspiration, not a result.","securityPrivacyImplications":"Officials say data is encrypted at rest and in transit, teams see only data needed for their work, and anonymized data is used where possible; no independent privacy review is described.","caveats":"Outcome figures come from city officials, not an independent evaluation; the welfare-uptake gain is unquantified; the attendance gain is a city hall claim; pilot sample was 300 women with no control group.","streamIds":["health-human-services"],"roles":{"sales":"The problem Rio is solving is residents who are eligible but never reached. For U.S. counties, the conversation is about proactive outreach on integrated data, starting with one narrow program such as postpartum follow-up. Lead discovery with data-sharing readiness and consent rules, and avoid promising maternal-mortality reductions, which the source presents only as an official's expectation.","engineering":"This is a data integration story first: AI outreach works because health, social services and education data were linked into one lake with role-based views. Any U.S. analogue needs a master person index, consent and data-use agreements, and a messaging channel residents actually answer. Validate referral precision and contact rates before expanding, since phone changes and unknown-number avoidance were reported limits.","delivery":"Delivery means standing up nurse or caseworker teams who act on flags, with dashboards trimmed to action-relevant fields. Start with one cohort, track contact, response and referral rates, and expand in stages as Rio did. Assign a data-governance owner for cross-program access and an operations owner for follow-up capacity, since outreach without staff to respond creates new risk."},"roleDetails":{"sales":{"customerProblem":"Eligible residents fall through gaps between fragmented public services and are not reached until problems become serious.","stakeholders":["County health and human services directors","Maternal and child health program leads","Agency chief data officers"],"discoveryQuestions":["Which cross-program data can you legally link today?","Which channel do your residents reliably answer?","Who would act on an outreach flag within 24 hours?"],"valueHypothesis":"Linked data plus targeted outreach on one cohort could raise service uptake with existing staff, if follow-up capacity is planned.","potentialEngagement":"Data-readiness review and a single-cohort outreach pilot design (for example postpartum Medicaid members).","unsupportedClaims":["Maternal mortality will drop 30-35%","Results transfer directly to U.S. counties","Privacy controls were independently verified"]},"engineering":{"fit":"Proactive outreach and triage layer on top of an integrated cross-program data platform; not a clinical decision tool.","architectureIntegration":"Cross-program data lake, cohort identification rules or models, messaging channel with daily check-ins, nurse dashboard and case record linking all touchpoints.","prerequisites":["Master person index across programs","Data-sharing agreements and consent basis","Staffed follow-up team"],"constraints":["HIPAA, 42 CFR Part 2 and state privacy laws limit U.S. cross-program linkage","Messaging channel adoption differs by population","No independent outcome evaluation yet"],"security":"Role-based access, encryption in transit and at rest, minimization in dashboards, and audit logging of who viewed linked records.","proofOfValue":["Contact and response rates for the target cohort","Share of flags that lead to a confirmed need","Staff time per resolved referral"]},"delivery":{"work":"Define the cohort, link required data, configure outreach scripts and escalation, train nurses or caseworkers, and report weekly on contact and referral outcomes.","dependencies":["Data-sharing agreements across departments","Messaging platform contract","Clinical or casework staffing for follow-up"],"ownership":"County health department program lead, with the agency chief data officer owning data access.","skillsAdoption":"Staff must learn remote risk assessment from self-reported information, which Rio nurses had to develop.","governanceCheckpoints":["Privacy and legal review of each data link","Monthly review of missed or false flags","Equity review of who is and is not reached"],"acceptanceCriteria":["Target cohort contact rate measured and reported (proposed)","Every red-flag response handled by a named clinician within a set time (proposed)","No data field shown without a documented use (proposed)"],"risks":["Outreach volume exceeding follow-up capacity","Privacy breach from a broad data lake","Missing residents who change phones or ignore unknown numbers"]}},"retrievedAt":"2026-11-10T22:19:20-06:00","enrichedAt":"2026-11-10T22:22:56-06:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Reaches residents who cannot travel or navigate digital government, but excludes those without phones; nurses take on remote assessment work.","procurementImplications":"Built by the city technology company Iplan Rio and funded in part by a Bloomberg Philanthropies Mayors Challenge award rather than a vendor procurement.","operatingModelImplications":"Moves agencies from waiting for residents to seeking them out, which requires standing follow-up capacity across departments.","peerReview":"unknown","sourceVerification":{"openedUrl":"https://www.jmir.org/2026/1/e115514","referenceExcerpt":"In a two-month pilot with 300 women, 15 reported significant warning signs and were referred back to maternity services","promptVersion":"sled-research-v4.0 HHS nightly brief","model":null,"basis":"agent-reported inspection"}}}]}