{"resourceId":"naco-911-behavioral-health-series-lyon-county-betty-ai-35pct-calls-20261028","versions":[{"version":"external-f0dbbab6381472ee0ca7a79760ef1df3f0f0365ac05548e1c8fe894742dfdd41","resource":{"id":"naco-911-behavioral-health-series-lyon-county-betty-ai-35pct-calls-20261028","title":"NACo 911 workforce brief: Lyon County, Kan., AI assistant 'Betty' now handles about 35% of calls after a four-month pilot, with staff monitoring every AI interaction","organization":"National Association of Counties (NACo); Lyon County Emergency Communications Center (Kansas)","sector":"911 emergency communications and telecommunicator workforce","geography":"United States (county case studies incl. Lyon County, Kansas)","publishedAt":"2026-10-28","publicationDate":"2026-10-28","eventDate":null,"sourceName":"National Association of Counties","sourceLabel":"NACo Series: 911 and Behavioral Health","sourceUrl":"https://www.naco.org/resource/naco-series-911-and-behavioral-health","evidenceClass":"public-sector-association","outcomeClass":"effective","topics":["accessibility-workforce","operating-model","knowledge-work"],"finding":"NACo's two-part 911 and behavioral health series (Oct. 28) profiles county call-handling and telecommunicator workforce practices. It states PSAPs handle more than 240 million 911 calls a year, including an estimated 24 million behavioral health calls, staffed by more than 100,000 telecommunicators, and that 68% report daily pre-shift and persistent stress. In its Lyon County, Kan., case, the emergency communications center introduced an AI assistant, 'Betty', in December 2025 for routine non-emergency calls (about 80% of volume) amid staffing shortages. After a four-month pilot it manages roughly 35% of calls; emergency calls stay human-handled and staff monitor all AI interactions. The center reports lower workload pressure, better morale, time to train two new telecommunicators, and commissioners redirected an unfilled position's funds to pay for Betty.","sledRelevance":"A national county association now presents a non-emergency AI call assistant as a workforce strategy, giving county leaders a peer case with a stated share of calls and a staffing-budget trade-off.","evidence":"Association case summary based on county-reported outcomes. The 35% figure has no stated denominator (all calls or non-emergency calls), period or error data; morale and workload improvements are self-reported without measurement. Vendor and cost are not given.","architectureImplications":"AI sits on the non-emergency path only, with emergency calls routed to humans and every AI interaction monitored by staff.","governanceImplications":"Shows a human-monitoring control and an explicit budget decision (repurposing an unfilled position) that county boards should document and revisit.","securityPrivacyImplications":"Source does not describe call recording, transcript retention or vendor data access for AI-handled calls.","caveats":"Brief case vignette within a broader behavioral health series; figures reported by the center via NACo, not independently evaluated; 'roughly 35% of calls' is ambiguous about denominator; no misroute or caller-outcome data.","streamIds":["emergency-services"],"roles":{"sales":"For sales, the framing point is that NACo, a body county leaders trust, now cites Lyon County's AI assistant handling roughly 35% of calls as a workforce measure, paid for by an unfilled position. Conversations with county boards can start from staffing stress rather than technology. A bounded engagement is a workload and call-mix analysis before any AI pilot. The brief cannot support claims about accuracy, caller satisfaction or budget savings.","engineering":"For pre-sales engineering, the design is AI on the non-emergency line with emergency calls kept human and every AI interaction monitored. That monitoring load must be sized. Prerequisites are call-type classification, transfer paths and logging. Constraints include an undefined 35% denominator and a single small center. Proof of value should establish the baseline non-emergency share, measure monitored interactions per shift and audit a sample of AI-handled calls.","delivery":"For delivery, the work is a staged pilot like Lyon County's four months, with staff monitoring built into shifts and a decision point on funding. Ownership sits with the emergency communications director and county commission. Acceptance should set a measured share of calls handled, a misroute threshold and a staff workload survey. Risks include AI substituting for hiring, unmeasured errors and monitoring fatigue."},"roleDetails":{"sales":{"customerProblem":"Small county 911 centers face staffing shortages, a thin applicant pool and high telecommunicator stress while routine non-emergency calls dominate volume.","stakeholders":["Emergency communications director","Telecommunicators","County commissioners","County HR and budget office","AI vendor"],"discoveryQuestions":["What share of your call volume is routine non-emergency?","Would you fund AI from a vacant position, and how would you revisit that?","Who would monitor AI interactions and how much time would that take?"],"valueHypothesis":"An AI assistant on non-emergency calls may free telecommunicator time for training and emergencies, as Lyon County reports, if monitoring is sustainable (interpretation).","potentialEngagement":"A call-mix and workload baseline study with a four-month pilot design and evaluation plan.","unsupportedClaims":["Accuracy of AI-handled calls","Cost savings","Caller satisfaction","Transferability to large centers"]},"engineering":{"fit":"Fits small and mid-size PSAPs where routine non-emergency calls are most of the volume.","architectureIntegration":"AI assistant on the non-emergency line with human handling of emergency calls and staff monitoring of every AI interaction.","prerequisites":["Call-type data to size the non-emergency share","Transfer paths to telecommunicators","Interaction logging for monitoring"],"constraints":["35% denominator undefined","Single small county","Vendor not named"],"security":"Define retention and vendor access for AI call recordings and transcripts.","proofOfValue":["Baseline non-emergency share before launch","Monitored interactions per shift","Audited sample of AI-handled calls for errors"]},"delivery":{"work":"Run a staged pilot with staff monitoring, then a funding and scope decision.","dependencies":["Vendor configuration","County commission budget approval","Telecommunicator monitoring time"],"ownership":"Emergency communications director, accountable to the county commission.","skillsAdoption":"Telecommunicators learn to monitor and correct AI interactions; supervisors learn to review AI logs.","governanceCheckpoints":["End-of-pilot review before funding","Written rule that emergency calls stay human-handled","Periodic review of position-funding trade-off"],"acceptanceCriteria":["Proposed: measured share of non-emergency calls handled with a stated denominator","Proposed: audited misroute rate below an agreed threshold","Proposed: staff workload survey shows no increase"],"risks":["AI used to avoid hiring","Unmeasured errors","Monitoring fatigue","Vendor dependence"]}},"retrievedAt":"2026-10-30T21:57:51-05:00","enrichedAt":"2026-10-30T22:02:41-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Frames AI as relief for a small, stressed workforce: freed time went to training new telecommunicators rather than cutting staff, though funding came from an unfilled position.","procurementImplications":"Funding the tool from a vacant position is a budget precedent; boards should weigh whether it substitutes for hiring.","operatingModelImplications":"Telecommunicators shift from answering routine calls to monitoring AI interactions and handling emergencies.","sourceVerification":{"responseId":"curl-naco-2026-10-30","openedUrl":"https://www.naco.org/resource/naco-series-911-and-behavioral-health","referenceExcerpt":"After a four-month pilot, Betty now manages roughly 35% of calls, while emergency calls stay human-handled and staff monitor all AI interactions.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}