State & Local GovernmentIssue 382 records
State Government
California's Department of General Services launched Ask Ozzie, a scoped chatbot for small and disabled-veteran business certification questions with staff escalation, aiming to free staff for backlog work but publishing no baseline or testing. StateTech's Oct 15 feature has Maryland, Caltrans and Colorado officials on moving GenAI from pilot to production, including gated access and production budgeting; Maryland's headline numbers are unaudited, and its 97% chatbot figure was earlier reported as scenario accuracy rather than live resolution.
- Evidence records
- 2
- Cross-source patterns
- 1
- Evidence classes
- 1 independent reporting1 government evaluation
- Outcomes
- 2 emerging
- Source freshness
- 2 new this fortnight
- Research completed
- 2026-10-16
Choose a role to see its takeaway beside every record in the ledger.
Patterns across the evidence
Synthesis · Lighthouse Advisory interpretation
State service chatbots launch with human escalation but without published, consistently defined outcome metrics
Both items tonight describe state chatbots with staff handoff: California DGS's Ask Ozzie launched with goals (freeing staff for certification backlog) but no baseline or accuracy data, and Maryland's SUN Bucks chatbot is cited at 97% as a live resolution rate in StateTech after being reported in 2025 as scenario accuracy. Interpretation: buyers and oversight bodies cannot compare chatbot performance until agencies publish what they measure and how.
Operating questionBefore launch, which metrics will this agency publish for its chatbot (scenario accuracy, live resolution, escalation rate, repeat contact, backlog impact), and how are they defined?
Supporting evidenceMaryland (Senior AI Advisor), California State Transportation Agency/Caltrans, Colorado OIT (former CIO); reported by StateTechCalifornia Department of General Services (DGS), Procurement Division, Office of Small Business and Disabled Veteran Business Enterprise Services (OSDS)
Evidence ledger
Full record · every source keeps its link and limitations
StateTech (Oct 15): Maryland, Caltrans and Colorado officials describe moving GenAI from pilot to production, with Maryland citing ~55% employee adoption and a 97% SUN Bucks chatbot figure
StateTech's Oct 15 feature quotes state officials on scaling AI past pilots. Maryland's Senior AI Advisor says Gemini and Copilot rollouts have reached adoption by about 55% of employees, one agency saved about $400,000 by doing work in Gemini instead of a vendor contract, and the SUN Bucks nutrition-benefit chatbot has had over 30,000 chats and resolves the question in 97% of cases. Caltrans' information security officer says two GenAI efforts (Road Safety Insights, Traffic Mobility Insights) moved into production and advises planning for production funding, support, cybersecurity, licenses and monitoring when the pilot is planned. Colorado's former CIO says Gemini went from a 150-plus-user pilot that deliberately included employees with disabilities to 18 agencies and about 2,000 initial users, gated by request, manager approval and responsible-use training. The Center for Democracy and Technology analyst cites unclear use cases and lack of scale-up funding as stall causes.
Why it matters, evidence and limitations
- Why it matters
- Fresh, quotable peer evidence for the pilot-to-production problem most state CIOs now face, with concrete operating practices (access gating, mandatory training, production budgeting). Usefulness is limited because metrics are officials' verbal claims, and the Maryland chatbot figure is framed differently from the state's earlier published description.
- Evidence and measured results
- Opened the StateTech article over HTTPS (curl, HTML text extraction) and read all state sections. For scrutiny, opened GovTech's Oct 29, 2025 Maryland report and Code for America's Jul 31, 2025 note on the Maryland Summer EBT chatbot. GovTech (2025) reported ~43,000 employees with Gemini access, ~12,500 active users, about 30% regular use, the same $400,000 Maryland Department of the Environment saving, and SUN Bucks Chat 'accurately answered questions in 97 percent of scenarios' with 16,000-plus users.
- Limitations and uncertainty
- Vendor-adjacent outlet (StateTech is published by CDW, an IT reseller); claims are officials' quotes, not audited results. The Maryland 97% figure appears here as a resolution rate across 30,000-plus chats, whereas GovTech in Oct 2025 reported it as accuracy across test scenarios by the state benefits team; the two are different measures and the methodology is unpublished. The '55% adoption' is not defined (access, monthly use, or regular use) and compares with about 30% regular use reported in Oct 2025. The $400,000 saving is the same example reported a year ago, not new. Caltrans production tools were archived earlier via NASCIO award coverage; only the pilot-to-production advice is new.
California DGS launches 'Ask Ozzie' AI chatbot for small business and disabled veteran certification questions, with staff escalation and a backlog-reduction goal
On Oct 14, 2026, California DGS launched Ask Ozzie, an AI chatbot on the OSDS webpage that answers common Small Business, Disabled Veteran Business Enterprise and nonprofit certification questions 24/7. DGS says OSDS fields thousands of questions annually and more than two-thirds of its calls and emails involve routine certification topics. Ask Ozzie is built on a curated knowledge base of the most frequent certification questions, answers only what it can confidently address, routes topics needing human expertise to the right place, lets users connect with staff at any point, and gives tailored contacts for OSDS or the Cal eProcure team at FI$Cal. The stated goal is to free staff time for certification file review, reducing backlog and improving turnaround.
Why it matters, evidence and limitations
- Why it matters
- A concrete example of a state procurement office using AI to deflect routine supplier inquiries so staff can work certification backlogs, a pattern transferable to licensing, permitting and vendor-registration offices. Relevance is limited because DGS published no baseline volumes, accuracy testing, model or vendor details, or success measures.
- Evidence and measured results
- Opened the DGS press release and the OSDS program page over HTTPS (curl, HTML text extraction); the OSDS page shows the Ask Ozzie entry point. The release states the Oct 14, 2026 publication date. No independent evaluation or metrics exist yet.
- Limitations and uncertainty
- Agency announcement on launch day; benefits (staff time freed, backlog reduction, faster turnaround) are goals, not results. 'Thousands of questions annually' and 'more than two-thirds' are unquantified operator statements. No vendor, model, cost or evaluation is named. Whether it uses generative AI is not stated.
How to read this edition
Source findings, measured results and limitations come from the cited publications. Patterns, operating questions, role takeaways and implementation considerations are Lighthouse Advisory interpretation, stated as questions to validate locally rather than guaranteed outcomes. Vendor and operator claims are labeled as claims. Full research method.
- Independent reporting
- Independent reporting with attributable sources but without a formal evaluation design.
- Government evaluation
- A public body’s measured evaluation or documented pilot.