{"resourceId":"tennessee-iris-professional-licensing-chatbot-20260922","versions":[{"version":"external-1ec9c0c5322d905292f6934b3f3d8178a9afa587caa34a55705ab91978bfa86c","resource":{"id":"tennessee-iris-professional-licensing-chatbot-20260922","title":"Tennessee launches Iris AI chatbot for 357,000 professional licensees via CORE","organization":"Tennessee Department of Commerce & Insurance (TDCI)","sector":"State professional licensing / regulatory boards customer service","geography":"Tennessee, United States","publishedAt":"2026-09-22","publicationDate":"2026-09-22","eventDate":"2026-09-22","sourceName":"WZTV Fox 17 — Meet Iris: Tennessee launches AI chatbot for professional licensing help","sourceLabel":"Independent local reporting of state agency chatbot launch announcement","sourceUrl":"https://fox17.com/news/local/tennessee-launches-ai-chatbot-for-professional-licensing-help","evidenceClass":"independent-reporting","outcomeClass":"emerging","topics":["knowledge-work","accessibility-workforce","operating-model","data-security"],"finding":"Why now: same-day (2026-09-22) TDCI announcement of Iris, an English/Spanish 24/7 licensing chatbot on CORE and the Regulatory Boards site for 26 programs covering more than 357,000 licensees. Soft-launched in August; broader announcement Tuesday. Agency says Iris can access applications/files in CORE but cannot share customers' personal data, and will escalate when it cannot answer. 2025 CSC volume (163,910 applications; 220,960 inquiries; >4-minute wait-time cut; 96% satisfaction) is pre-Iris baseline context, not an Iris outcome metric.","sledRelevance":"Direct state-government citizen-service AI for licensing operations—high-volume shared customer-service channel transferable to other state commerce/licensing agencies, with explicit human handoff and PII non-disclosure claims that still need independent verification.","evidence":"Fox 17 attributes facts to TDCI announcement and Commissioner Carter Lawrence quotation. Soft launch August; public announcement 2026-09-22. No Iris-specific deflection, FCR, or error-rate metrics published in the opened article.","architectureImplications":"Chat embedded on CORE licensing portal and Regulatory Boards web property; bilingual routing; access to licensee application/file data in CORE with asserted personal-data non-disclosure; escalation path to human CSC (8 a.m.–4:30 p.m. CT weekdays) for unanswered questions.","governanceImplications":"Agency-framed as seamless transition between human CSC and AI; optional after-hours coverage. No published evaluation design, logging/audit policy, or independent accuracy audit in this article.","securityPrivacyImplications":"TDCI asserts Iris can access CORE application/file data but cannot share customers' personal data—buyers should verify access controls, retention, and prompt/log redaction independently.","caveats":"No measured Iris outcome metrics yet; 2025 CSC wait-time and satisfaction figures predate Iris and must not be attributed to the chatbot. Single local-news source; official tn.gov press release not located during this run.","streamIds":["state-government"],"roles":{"sales":"Sales framing: Tennessee just productized a licensing-desk AI for 357,000 licensees without publishing Iris outcome KPIs yet. Lead with soft-launch discipline and PII non-disclosure claims, then ask how the agency will prove accuracy across 26 boards. Bound any proposal to a 4–6 week readiness review—content inventory, escalation UX, bilingual coverage, and CORE access controls—before statewide expansion talk. Do not sell guaranteed call deflection from 2025 CSC baselines.","engineering":"Engineering framing: Iris sits on CORE plus Regulatory Boards web, reads application/file context, and must refuse personal-data disclosure while escalating unknowns. Validate identity/session boundaries between anonymous FAQ and authenticated licensee views, Spanish parity, and audit logs for retrievals. Proof-of-value should measure answerability, escalation rate, and hallucination spot-checks on a stratified sample of the 26 programs—not wait-time claims from the prior year.","delivery":"Delivery framing: stand up content owners per regulatory board, CSC escalation scripts, and after-hours monitoring before scaling marketing of Iris. Acceptance criteria should include documented PII non-disclosure tests, bilingual regression packs, and a human-handoff SLA during CSC hours. Primary risk is silent wrong answers on licensing eligibility that licensees treat as official guidance without remediation owners."},"roleDetails":{"sales":{"customerProblem":"State licensing agencies face high inquiry volume outside business hours and need consistent answers across many board programs without promising unproven deflection.","stakeholders":["TDCI commissioner / CSC leadership","Regulatory board program owners","CORE platform owner","State privacy/security officer"],"discoveryQuestions":["What Iris metrics will replace 2025 CSC baselines after 90 days?","How is CORE data scoped for the bot versus human agents?","Who owns wrong-answer remediation per board?","What is the Spanish content parity process?"],"valueHypothesis":"A bounded readiness and measurement engagement can convert a soft-launched licensing bot into an evidence-backed channel without claiming pre-Iris wait-time wins.","potentialEngagement":"4–6 week Iris measurement and escalation-design review across a sample of high-volume boards.","unsupportedClaims":["Guaranteed call-volume reduction from Iris","Equating 2025 CSC satisfaction/wait-time to Iris performance","Claiming zero PII risk without independent control testing"]},"engineering":{"fit":"Fit for FAQ/navigation and status-style licensing questions with mandatory human escalation; unfit as sole authority for eligibility determinations.","architectureIntegration":"Embed on CORE and Regulatory Boards sites; retrieve from approved licensing content and scoped application metadata; bilingual NLU; escalate to CSC.","prerequisites":["Allowlisted content corpus per board","CORE API/auth boundaries","Escalation ticket/chat bridge","PII redaction and access logging"],"constraints":["Tennessee-only regulatory rules","Weekday CSC coverage window","No vendor named in source"],"security":"Verify claims that Iris accesses files yet cannot share personal data; test prompt-injection and cross-licensee leakage.","proofOfValue":["Stratified accuracy sample across 26 programs","Escalation-rate and unanswered-rate dashboards","Spanish parity spot checks","After-hours vs in-hours quality comparison"]},"delivery":{"work":"Content normalization, escalation playbooks, monitoring, and board-owner training for Iris operations.","dependencies":["CORE platform team","Each regulatory board content owner","CSC staffing for handoffs","Privacy review of access patterns"],"ownership":"TDCI Customer Service Center with board program co-owners for content.","skillsAdoption":"CSC agents need bot-oversight skills; board staff need FAQ curation habits.","governanceCheckpoints":["Pre-expansion accuracy gate","Monthly wrong-answer review","Privacy access audit","Change control when statutes/rules update"],"acceptanceCriteria":["Documented escalation when Iris cannot answer","PII non-disclosure test cases pass","Published Iris-specific KPIs within 90 days","Bilingual critical-path scripts approved"],"risks":["Licensees treat incorrect bot guidance as binding","Uneven quality across 26 programs","Scope creep into transactional filing without identity proofing"]}},"retrievedAt":"2026-09-22T22:15:48-05:00","enrichedAt":"2026-09-22T22:28:29-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"English and Spanish support and after-hours availability expand access relative to weekday CSC hours; workforce implication is CSC triage plus content ownership for licensing FAQs rather than replacement of licensed-profession review.","procurementImplications":"Article does not name the chatbot vendor or contract vehicle; discovery should identify platform, data-processing terms, and CORE integration ownership.","operatingModelImplications":"Soft-launch-then-announce pattern; human CSC remains primary for complex cases; success depends on clear escalation UX and content freshness across 26 regulatory programs.","sourceVerification":{"openedUrl":"https://fox17.com/news/local/tennessee-launches-ai-chatbot-for-professional-licensing-help","referenceExcerpt":"Iris is designed to assist the more than 357,000 licensees covered by the division’s 26 regulatory programs","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}