{"resourceId":"ohio-das-ai-cloud-modernization-nascio-2026","versions":[{"version":"external-4588039a61d6d8fd04513b625be2894a9f63810baa2efcad1bbd173a71967e53","resource":{"id":"ohio-das-ai-cloud-modernization-nascio-2026","title":"Ohio DAS NASCIO entry: AI modernization cites 50–60% doc/estimation cut, 40% faster analysis","organization":"Ohio Department of Administrative Services — Cloud Center of Excellence (CCoE)","sector":"State shared services / legacy modernization / cloud center of excellence","geography":"Ohio executive-branch agencies via DAS CCoE (project dates May 2025–April 2026)","publishedAt":"2026-08","publicationDate":"2026-08-01","eventDate":"2026-09-30","sourceName":"NASCIO State IT Recognition Awards library (nomination PDF)","sourceLabel":"Driving Operational Excellence: How AI Accelerated Statewide Cloud Modernization","sourceUrl":"https://www.nascio.org/wp-content/uploads/2026/08/OH_Operational-Efficiency.pdf","evidenceClass":"public-sector-association","outcomeClass":"effective","topics":["developers-agents","infrastructure","operating-model","governance-procurement"],"finding":"Ohio DAS CCoE’s NASCIO Operational Efficiency nomination (project May 2025–April 2026) describes embedding AI-assisted techniques into early modernization phases—system interpretation, functional summarization, dependency mapping, and specification generation—using two methodologies (context engineering for deep undocumented legacy; deterministic agent-based execution for structured upgrades) under unified governance. Claimed results include 50–60% reduction in documentation/estimation/backlog time, 40% faster API specification/system analysis, 20–40% code-generation throughput improvement, and 50% reduction in code-review/debugging time across pilot engagements, with human judgment retained and AI kept inside state cloud/security controls and inventory/oversight requirements.","sledRelevance":"Primary association evidence with quantitative modernization metrics matching tonight’s GovTech award coverage. Why now: NASCIO announced recipients 2026-09-30 (Ohio Operational Efficiency winner per GovTech); PDF not previously archived; newly relevant as the underlying measured claim behind same-day award press.","evidence":"Downloaded and pdftotext’d the NASCIO nomination PDF via curl; confirmed project dates, dual-methodology framing, responsible-AI governance language (human review, inventory/oversight), and Impact metrics (50–60%, 40%, 20–40%, 50% review/debug).","architectureImplications":"Separate AI discovery/context-engineering from deterministic agent execution; produce human-accepted blueprints before AI-assisted coding; keep tooling inside state cloud/security boundaries.","governanceImplications":"Ohio’s pitch treats governance as co-equal with tech—mandatory human review, hallucination-risk controls, enterprise AI inventory—buyers should require the same artifacts before scaling modernization AI.","securityPrivacyImplications":"Legacy-code AI analysis can surface secrets and PII in old systems; sandbox corpora, redact secrets, and log model access to source.","caveats":"Self-reported nomination for an awards program; pilots may not represent all agencies; percentages lack independent audit methodology detail; eventDate reflects award announcement relevance, not PDF authorship date.","streamIds":["state-government"],"roles":{"sales":"Frame Ohio’s NASCIO modernization entry as a measured shared-services pattern—AI compresses documentation and analysis before coding, with claimed 50–60%/40% cycle cuts—not a turnkey agent that modernizes alone. Discovery should ask which legacy class (undocumented versus structured upgrade) dominates and what human acceptance gates exist. Bounded engagement: a CCoE-style discovery pilot on one undocumented system with timed baselines. Unsupported claims include guaranteed statewide 50–60% savings, that agentic execution is safe without deterministic scopes, or that nomination metrics equal audited ROI.","engineering":"Fit is AI-assisted legacy interpretation plus optional deterministic code-assist inside state cloud controls, always with human-owned specs. Architecture: sandboxed code corpora to context-engineering summaries to SME blueprint sign-off to constrained generation to inventory logging. Prerequisites: CCoE or equivalent, classification, baseline metrics. Constraints: nomination self-report; dual-method complexity. Proof of value is before/after documentation and analysis hours on a named application plus defect/rework rates.","delivery":"Delivery stands up a modernization AI playbook, trains analysts on review discipline, and runs one agency engagement with published KPIs. Dependencies: DAS/CCoE, agency app owners, security, budget for shared investment. Ownership of accepted system definition remains with the agency. Acceptance criteria: signed blueprint, measured cycle-time delta, AI use logged in inventory. Risks: hallucinated dependencies, secret leakage from legacy code, and scaling methods before governance artifacts exist."},"roleDetails":{"sales":{"customerProblem":"Legacy modernization stalls on undocumented systems; states need faster discovery without losing human ownership.","stakeholders":["DAS/CCoE","Agency application owners","State CIO","Security","Budget/shared-services finance"],"discoveryQuestions":["Which apps lack documentation?","What baseline hours exist for discovery?","Who signs AI-produced specs?","Are code-AI tools already inventoried?"],"valueHypothesis":"AI-assisted discovery under CCoE governance can cut documentation/analysis time if SME acceptance and sandboxes are mandatory.","potentialEngagement":"Single-application AI discovery pilot with timed baselines and playbook draft.","unsupportedClaims":["Guaranteed 50–60% statewide savings","AI replaces modernization SMEs","Nomination metrics are independently audited","Agentic coding needs no deterministic scope"]},"engineering":{"fit":"Sandbox AI for legacy interpretation and constrained generation after human blueprint approval.","architectureIntegration":"Classified code corpus → context-engineering → SME sign-off → deterministic/agent assist → CI with human review.","prerequisites":["Cloud/security controls","Inventory process","SME owners","Secret scanning"],"constraints":["Self-reported pilots","Heterogeneous legacy stacks","Hallucination risk on undocumented logic"],"security":"No production credentials in prompts; redact secrets; log model access to source.","proofOfValue":["Doc/estimation hours before/after","Analysis cycle time","Rework defects post-AI assist"]},"delivery":{"work":"Pilot one system; capture metrics; publish playbook; train agency team for reuse.","dependencies":["CCoE","Agency owner","Security review","Procurement for tools"],"ownership":"Agency owns accepted system definition; CCoE owns playbook quality.","skillsAdoption":"Analyst training on validating AI summaries and dependency maps.","governanceCheckpoints":["Sandbox approval","Blueprint sign-off","Inventory update","Pilot retrospective"],"acceptanceCriteria":["Measured cycle-time delta","Signed blueprint","No unsupervised production deploy"],"risks":["Secret leakage","Metric overclaim","Method mismatch to legacy class"]}},"retrievedAt":"2026-09-30T22:22:05-05:00","enrichedAt":"2026-09-30T22:22:05-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Program claims agencies gain skills to continue modernization independently—plan knowledge transfer so AI-accelerated docs remain usable by non-vendor staff including accessibility/ops reviewers.","procurementImplications":"Shared-investment/CCoE model distributes cost; contracts for code-AI should require audit logs, no training on agency code without approval, and human acceptance criteria for deliverables.","operatingModelImplications":"CCoE-led playbook with agency ownership after enablement is the operating model; AI accelerates pre-build understanding rather than replacing agency decision rights.","sourceVerification":{"responseId":"curl-pdftotext-ohio-oe-20260930T222205","openedUrl":"https://www.nascio.org/wp-content/uploads/2026/08/OH_Operational-Efficiency.pdf","referenceExcerpt":"To date, the effort has resulted in a 50–60% reduction in documentation and estimation time","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}