{"resourceId":"texas-treo-sam-regulatory-chatbot-statescoop-20260515","versions":[{"version":"external-0ba739ec74be2654fd543a25a17954f80a7dbf5d3c58ea2d19e1c1fdb926e685","resource":{"id":"texas-treo-sam-regulatory-chatbot-statescoop-20260515","title":"Texas TREO launches SAM regulatory chatbot with citation-backed licensing answers (~15 seconds claimed)","organization":"Texas Regulatory Efficiency Office (Office of the Governor); Vulcan Technologies (vendor)","sector":"State regulatory navigation / professional licensing information / administrative-code modernization","geography":"Texas, United States","publishedAt":"2026-05-15","publicationDate":"2026-05-15","eventDate":"2026-05-15","sourceName":"StateScoop — Texas efficiency office launches AI chatbot to help public navigate regulations, licensing","sourceLabel":"Independent reporting on Efficiency.Texas.Gov / SAM launch with TREO director quotes and disclosure details","sourceUrl":"https://statescoop.com/texas-efficiency-office-launches-ai-chatbot-to-help-public-navigate-regulations-licensing/","evidenceClass":"independent-reporting","outcomeClass":"mixed","topics":["knowledge-work","governance-procurement","accessibility-workforce","operating-model","data-security"],"finding":"Why now: NASCIO Annual Conference week (Sep 27–30 San Diego) and Sep 27 Forbes commentary resurfaced TREO's SAM as a statewide regulatory front door while Digital States Texas coverage stressed community enablement without this product deep-dive; still unarchived. StateScoop (2026-05-15) reports Efficiency.Texas.Gov's SAM chatbot answers rules/licenses/permits in English/Spanish from statutes and the Administrative Code, with informational-not-legal-advice disclaimers, public-records privacy warnings, and pre-deployment evaluation for accuracy, citation fidelity, and refusal behavior. TREO separately cites >435 regulations for amendment/repeal (~69,000 words; $123M estimated taxpayer savings). Director Greener describes multi-model orchestration and ~15-second answers for questions like becoming an electrician; stress-test anecdote of 25,000 concurrent agents. Independent reporting notes NYC chatbot failure precedent as caution.","sledRelevance":"Public-facing, citation-grounded state regulatory/licensing assistant with explicit refusal/disclaimer design—directly transferable pattern for licensing and administrative-code navigation (not eligibility adjudication).","evidence":"Opened StateScoop article via curl; corroborating NASCIO award narrative PDF (TX_ArtificialIntelligence.pdf) opened and text-extracted for architecture/metrics detail. $123M/435-regs figures are TREO/Governor claims, not independent fiscal audits in these sources.","architectureImplications":"Statute/TAC-grounded retrieval with multi-model agent orchestration (Vulcan); bilingual UX; Regulation Evaluation Portal for public feedback; AI-assisted internal code review feeding public dashboard of recommended changes.","governanceImplications":"SB 14 statutory TREO mandate and funding; AI disclosure that user is not speaking to a person; not-legal-advice framing; advisory navigation only.","securityPrivacyImplications":"Warns users not to enter sensitive personal information because inputs may be public records; accuracy/citation/refusal evaluation claimed pre-launch without published error rates.","caveats":"May 15 launch vintage—included for NASCIO-week relevance and archive gap, not as brand-new publication. $123M savings and 15-second claims are operator/vendor-attributed; NYC MyCity chatbot failures cited as cautionary parallel. Not legal advice and not a substitute for agency determination.","streamIds":["state-government"],"roles":{"sales":"Sales framing: Texas TREO's SAM—resurfaced this NASCIO conference week—shows a citation-backed public regulatory/licensing chatbot with bilingual answers, refusal behavior, and ~15-second electrician-licensing examples, plus a parallel 435-rule modernization ledger. Open discovery on citation fidelity sampling and how $123M savings were estimated. Bound offers to grounded administrative-code navigation and modernization analytics—not unsupervised license grants or legal advice.","engineering":"Engineering framing: design SAM-class assistants as statute/TAC-grounded retrieval with multi-model orchestration, bilingual UX, disclosure gates, and hard refusal for off-policy prompts. Proof-of-value measures citation fidelity, refusal correctness, and time-to-answer on a licensed-profession gold set—not the 25,000-agent stress anecdote alone. Keep determinations and fee payments outside chat completion; assume prompts may be public records. Publish eval harness results before scaling beyond the first board cluster.","delivery":"Delivery framing: stand up corpus refresh for statutes/TAC, citation UX, privacy banners, agency escalation links, and a modernization recommendation dashboard with TREO and board owners. Acceptance criteria publish accuracy/refusal test results and keep chatbot KPIs separate from the $123M repeal estimate. Risks include hallucinated fees, users treating advisory chat as legal advice, public-records privacy surprises, and marketing that conflates rule-repeal savings with chatbot ROI."},"roleDetails":{"sales":{"customerProblem":"Licensing and regulatory agencies need a public front door that reduces multi-site research without creating unauthorized practice-of-law or wrong-fee harm.","stakeholders":["Governor efficiency office / TREO","Licensing boards","State CIO/DIR","Privacy/counsel","Vendor (Vulcan)"],"discoveryQuestions":["What gold-set citation fidelity rate is monitored in production?","How are TAC updates retrained?","What happens when SAM is unsure?","Is the $123M figure tied to chatbot use or separate rule reviews?"],"valueHypothesis":"A grounded regulatory-navigation engagement can replicate SAM's citation/disclaimer pattern with measurable fidelity and clear separation from modernization savings claims.","potentialEngagement":"6–8 week licensing FAQ corpus + citation QA + disclosure UX pilot for one board cluster.","unsupportedClaims":["Audited $123M chatbot ROI","Legal advice substitute","Unsupervised license issuance","Guaranteed 15-second answers statewide"]},"engineering":{"fit":"Strong fit for retrieval-grounded public regulatory assistants with bilingual needs and refusal policies.","architectureIntegration":"Primary statute/TAC knowledge base; multi-model orchestration for complex queries; disclosure/privacy interstitial; agency deep links; optional internal review analytics twin.","prerequisites":["Authoritative code corpus with versioning","Profession taxonomy/fee tables","Refusal/eval harness","Public-records logging policy"],"constraints":["Texas SB 14/TREO context","Not legal advice boundary","Public-records exposure of prompts","Vendor agentic stack portability"],"security":"Minimize PII collection; treat chats as potentially public; cite sources; block off-domain creative asks.","proofOfValue":["Citation fidelity on gold set","Refusal tests (jokes/illegal asks)","Time-to-answer distribution","Human escalation rate"]},"delivery":{"work":"Corpus pipelines, bilingual UX, disclosure banners, board escalation paths, evaluation dashboards, modernization portal ops.","dependencies":["TREO/legal","Licensing agencies","Vendor platform","Communications"],"ownership":"TREO owns public platform; boards own authoritative determinations; vendor operates model stack under state contract.","skillsAdoption":"Board staff need citation-review literacy; analysts need code-modernization AI-assist skills.","governanceCheckpoints":["Pre-launch accuracy/refusal gate","Corpus update cadence","Monthly hallucination sample","Separation of chatbot KPIs from repeal savings"],"acceptanceCriteria":["Published eval summary","Working not-legal-advice + privacy notices","Agency contact in every license answer","No automated license grant"],"risks":["Wrong fee/table hallucination","Public-records privacy surprises","Savings attributed to chatbot without causal link"]}},"retrievedAt":"2026-09-28T22:35:00-05:00","enrichedAt":"2026-09-28T22:48:00-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"English/Spanish support improves access; screen-reader/accessibility of chat UX not detailed. Staff impact is primarily TREO analysts using AI for code review, not mass workforce copilots.","procurementImplications":"Vulcan Technologies named as builder; SB 14 funds TREO ($22.8M/5 years per NASCIO narrative). Confirm data-processing, model-routing, and public-records terms.","operatingModelImplications":"Small TREO team (about 8) plus advisory panel; public portal dual-uses citizen Q&A and transparency dashboard for repeal recommendations.","sourceVerification":{"openedUrl":"https://statescoop.com/texas-efficiency-office-launches-ai-chatbot-to-help-public-navigate-regulations-licensing/","referenceExcerpt":"evaluated against representative regulatory questions prior to deployment for accuracy, citation fidelity, and refusal behavior","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}