{"resourceId":"gmtoday-waukesha-ava-aurelian-psap-20261002","versions":[{"version":"external-d37cc357cd81f1553c7b3e1c6d59379cee1cf0519dc1898699ea3d1319655121","resource":{"id":"gmtoday-waukesha-ava-aurelian-psap-20261002","title":"GMToday: Waukesha County Ava AI has handled 91,892 calls with ~53% automated under dispatcher review","organization":"Waukesha County Communications Center / Waukesha Freeman (GMToday)","sector":"County PSAP / emergency communications","geography":"United States (Wisconsin)","publishedAt":"2026-10-02","publicationDate":"2026-10-02","eventDate":"2026-10-02","sourceName":"GMToday / Waukesha Freeman","sourceLabel":"Local independent report on Waukesha County Communications Center Ava/Aurelian AI call handling","sourceUrl":"https://www.gmtoday.com/the_freeman/news/waukesha-county-ai-dispatch/article_fa66c4db-ba4c-5756-ba98-3719fd5b294e.html","evidenceClass":"independent-reporting","outcomeClass":"emerging","topics":["operating-model","accessibility-workforce","infrastructure","governance-procurement"],"finding":"On October 2, 2026, the Waukesha Freeman (GMToday) reports the Waukesha County Communications Center went live January 21 with Aurelian’s Ava conversational AI for call handling. Operations Manager Chris Becker states Ava has handled 91,892 calls since go-live, with about 53% automated in some way (gathering information for dispatchers, transferring callers, or providing information). Dispatchers still review every Ava action and flag transcription/location/question-order issues for vendor iteration. Staffing is approved for 46 dispatchers; Becker frames Ava as helping existing staff answer 911 faster rather than replacing dispatchers, with a satisfaction target near 60% automation versus vendor estimates of 70–75% non-emergency reduction. Fail-safes escalate disabled-vehicle and other cues to live dispatchers; Becker emphasizes humans ultimately decide. Other Wisconsin agencies (La Crosse first, Dane live, Brown preparing, Ozaukee non-emergency) are adopting Ava.","sledRelevance":"Direct PSAP emergency-communications AI evidence with measured call volumes and automation share under human review—core emergency-services dispatch stream.","evidence":"Local journalism quoting the communications-center operations manager with specific call counts and automation percentage; not an independent outcome audit.","architectureImplications":"Conversational agent with call-type routes, text/info responses, inter-agency transfers, and continuous human review feedback into vendor route changes.","governanceImplications":"Public discussion of automation rates and peer-county adoption contrasts with undisclosed 911 AI elsewhere; labor/funding framing treats AI as gap-fill not replacement.","securityPrivacyImplications":"Article does not specify audio retention, vendor model-training use, or CJIS controls—procurement must add those gates.","caveats":"Single-agency local reporting; metrics agency-reported; emergency-answer-time benefits not measured here.","streamIds":["emergency-services"],"roles":{"sales":"Frame Waukesha’s Oct 2 report—Ava handling 91,892 calls with ~53% automated under dispatcher review—as a capacity story for non-emergency load, not autonomous 911. Discovery should map current non-emergency share, review SLAs, escalation fail-safes, and union/staffing politics. Bounded engagement: design HITL review metrics and route-improvement loops for a PSAP conversational-AI pilot. Unsupported claims include guaranteed 70–75% call reduction, faster 911 answer times proven here, or replacement of dispatchers.","engineering":"Treat Ava/Aurelian as a routed conversational agent on non-emergency and transfer paths with mandatory dispatcher review of every action. Prerequisites: call-type taxonomy, transfer destinations (e.g., municipal PD), transcription QA, and CAD-adjacent logging. Constraints: agency-reported automation rates; emergency escalation depends on configured cues. Proof of value: time-to-911-answer, false-automation/escalation miss rate, and dispatcher review workload—not vendor demo scripts alone.","delivery":"Stand up go-live with continuous dispatcher feedback into vendor route changes, published escalation SOPs (disabled vehicle, background distress cues), and staffing plans that keep 46-FTE-scale human capacity. Ownership: communications-center operations for acceptance; vendor for route fixes. Risks: silent miss of escalating emergencies, transcription/location errors, and treating automation % as an outcome instead of a workload proxy."},"roleDetails":{"sales":{"customerProblem":"PSAPs need to clear non-emergency volume so limited dispatchers can answer true 911 faster.","stakeholders":["PSAP/comms center ops","Dispatchers/union","County executives","Municipal PD transfer partners"],"discoveryQuestions":["What % of lines are non-emergency today?","Who reviews every AI action and how fast?","What cues force live transfer?","What 911 answer-time baseline exists?"],"valueHypothesis":"Conversational AI can reduce non-emergency load if escalation fail-safes and human review are instrumented—not if automation % is the only KPI.","potentialEngagement":"6–8 week PSAP readiness: route inventory, review SOP, and KPI dashboard for an Ava-like pilot.","unsupportedClaims":["Guaranteed 70–75% reduction","Proven faster 911 answers in this article","Dispatcher replacement"]},"engineering":{"fit":"Non-emergency conversational AI with human review and emergency escalation transfers.","architectureIntegration":"Caller → Ava routes → info/transfer/gather → dispatcher review; CAD/telephony integration implied not specified.","prerequisites":["Call taxonomy/routes","Transfer directories","Review tooling","Baseline answer-time metrics"],"constraints":["Agency-reported metrics","Not autonomous emergency decision-making per Becker","Vendor iteration dependency"],"security":"Article does not detail data retention/vendor access; require CJIS/telephony security review before deploy.","proofOfValue":["911 answer-time delta","Missed-escalation audits","Dispatcher review minutes per 100 AI actions"]},"delivery":{"work":"Pilot routes, train dispatchers on review/flags, negotiate vendor change SLA, publish caller-facing expectations.","dependencies":["Telephony vendor","Aurelian success team","Municipal transfer partners","Labor agreement"],"ownership":"Communications center operations owns accept/reject of AI actions; vendor owns route engineering.","skillsAdoption":"Train staff to talk naturally to conversational AI and to flag transcription/location errors systematically.","governanceCheckpoints":["Go-live criteria","Weekly miss/escalation review","Public disclosure decision"],"acceptanceCriteria":["100% AI actions reviewable","Documented escalation cues","Answer-time KPI defined"],"risks":["Missed emergency cues","Over-trust in automation %","Transfer partner gaps"]}},"retrievedAt":"2026-10-03T22:01:34-05:00","enrichedAt":"2026-10-03T22:01:34-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Positions AI as helping approved 46-dispatcher staffing cope with volume; conversational UX guidance for callers (talk normally, not one-word IVR habits).","procurementImplications":"Ongoing daily vendor collaboration and customer-success review of dispatcher flags implies a managed service relationship, not a one-time buy.","operatingModelImplications":"Non-emergency AI triage with human final authority and configured emergency escalation; peer Wisconsin PSAPs spreading the same pattern.","sourceVerification":{"responseId":"curl-gmtoday-waukesha-ava-20261003T220134","openedUrl":"https://www.gmtoday.com/the_freeman/news/waukesha-county-ai-dispatch/article_fa66c4db-ba4c-5756-ba98-3719fd5b294e.html","referenceExcerpt":"Since then, Ava has handled 91,892 calls. About 53% have been automated in some way","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}