{"resourceId":"riverside-rivy-chat-voice-metrics-202608","versions":[{"version":"external-3866bc7e50b4fc1ecb4a8dd537e9f0bef6c81163475ca4ae8a4567519b70794b","resource":{"id":"riverside-rivy-chat-voice-metrics-202608","title":"Riverside Rivy interview reports 98.7% chat resource-match and 9% 311 voice adoption","organization":"City of Riverside","sector":"Municipal 311 and resident information services","geography":"Riverside, California, United States","publishedAt":"August 13, 2026 AI Center for Government interview","publicationDate":"2026-08-13","eventDate":null,"sourceName":"Partnership for Public Service AI Center for Government","sourceLabel":"Public-sector association interview with Riverside deputy CIO","sourceUrl":"https://aicenterforgovernment.org/2026/08/13/ai-bright-spot-assisting-residents-with-an-ai-chatbot-in-riverside-california/","evidenceClass":"public-sector-association","outcomeClass":"mixed","topics":["knowledge-work","accessibility-workforce","operating-model","data-security","governance-procurement"],"finding":"Riverside’s Rivy chat and 311 voice assistants, trained only on public website data after citywide employee testing, report 98.7% chat resource-match accuracy, 9% voice adoption when offered, and 89.9% first-call resolution among voice users, with live-agent transfer available.","sledRelevance":"Complements Riverside’s September council AI working-group coverage with operational chatbot/voice metrics and public-data-only training—useful for cities balancing launch claims against low voice uptake and escalation design.","evidence":"Deputy CIO describes >100,000 annual 311 calls with nearly half informational; internal pilot ~2,700 employees; guardrails redirect homework queries to library services; ongoing intent/response monitoring. Metrics are self-reported in an association ‘bright spot’ interview.","architectureImplications":"Shared knowledge base across web chat and 311 voice; constrain training to public site content; preserve live transfer during business hours.","governanceImplications":"Pair high accuracy claims with published error definitions and independent retests; council strategy work should incorporate these KPIs and escalation SLAs.","securityPrivacyImplications":"Public-data-only training reduces sensitive-data exposure but does not eliminate logging/privacy issues; homework/off-topic guardrails show need for abuse controls.","caveats":"Association bright-spot framing; metrics lack independent verification; 9% voice adoption limits claimed 311 impact.","streamIds":["local-government"],"roles":{"sales":"Interpretation: When cities cite Rivy-like accuracy, discover the match definition, voice opt-in rate and independent testing plans. Engage innovation, 311, communications and council sponsors. Ask how public-data-only training is enforced and how transfers work after hours. A bounded engagement can replicate the employee soft-launch and define accuracy/escalation SLAs. Value is 24/7 information routing with humans for complex requests—not proven call-center replacement. Do not treat 98.7% as transferable without method disclosure.","engineering":"Interpretation: Fit is retrieval-constrained resident Q&A on public content across chat and IVR. Integrate with 311 telephony and website CMS ownership. Prerequisites: corpus governance, intent monitoring, transfer paths and abuse/homework guardrails. Constraints: low voluntary voice adoption; business-hours live agents. Security: no private record training; define conversation log retention. Proof of value needs published accuracy protocol plus adoption and escalation rates, not a single percentage.","delivery":"Interpretation: Deliver continuous content refresh and 311 escalation quality monitoring. Dependencies include department content owners, 311 supervisors and telephony vendors. Ownership: innovation/IT for the assistant; 311 for voice operations; executives for accessibility commitments. Train staff on soft-launch testing and residents that live help remains. Governance checkpoints before expanding beyond public data; acceptance requires documented transfer success and independent accuracy sampling. Risks include overstated 311 impact given 9% adoption."},"roleDetails":{"sales":{"customerProblem":"Large volumes of routine 311/info calls leave residents waiting while answers sit scattered on city websites.","stakeholders":["Chief innovation / IT","311 operations","Department content owners","Mayor/council","Residents"],"discoveryQuestions":["How is 98.7% resource-match defined and sampled?","Why is voice adoption only 9%?","What is after-hours escalation policy?","How is public-data-only training enforced with vendors?"],"valueHypothesis":"Public-data chat/voice agents may improve informational routing if accuracy methods are published and live help remains available.","potentialEngagement":"Soft-launch design with department testing and dual chat/voice KPIs before public expansion.","unsupportedClaims":["98.7% accuracy in other cities without testing","Voice AI will absorb most 311 volume at 9% adoption","Public-data training eliminates all privacy risk"]},"engineering":{"fit":"Public-content conversational assistants for web chat and 311 voice with human transfer.","architectureIntegration":"Shared curated corpus tied to city website; IVR option into same guardrailed model; no private-case retrieval.","prerequisites":["Website content governance","Employee soft-launch","Intent/quality monitoring","Live-agent handoff"],"constraints":["Low voice opt-in","Business-hours human coverage","Association-reported metrics only"],"security":"Block private data training; control logs; harden against off-topic/abuse prompts.","proofOfValue":["Published accuracy protocol","Adoption and escalation rates","Independent sample retest"]},"delivery":{"work":"Run Rivy with ongoing content updates, intent review and 311 transfer quality checks.","dependencies":["CMS owners","311 supervisors","Telephony platform","Communications"],"ownership":"Innovation/IT owns product; 311 owns voice operations; council sets accessibility expectations.","skillsAdoption":"Train departments for soft-launch testing; coach agents on AI-to-human handoffs.","governanceCheckpoints":["Before adding non-public data","After major website restructures","Quarterly independent accuracy sample"],"acceptanceCriteria":["Live transfer works in tested scenarios","Public-data constraint verified","KPIs include adoption not only accuracy"],"risks":["Bright-spot bias","Low voice uptake undercutting ROI stories","Residual wrong answers on high-stakes topics"]}},"retrievedAt":"2026-09-18T03:07:15Z","enrichedAt":"2026-09-18T03:07:15Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Voice and chat extend access but low voice adoption and business-hours live agents mean phone equity and disability accommodations still depend on human 311 capacity.","procurementImplications":"If platform components are vendor-hosted, contracts must enforce public-data-only training and handoff SLAs; association case studies are not RFPs.","operatingModelImplications":"Employee soft-launch and continuous monitoring are part of the operating model, not one-time go-live tasks.","sourceVerification":{"openedUrl":"https://aicenterforgovernment.org/2026/08/13/ai-bright-spot-assisting-residents-with-an-ai-chatbot-in-riverside-california/","referenceExcerpt":"the chat agent successfully matches a resident’s request to the correct city resource 98.7% of the time","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}