{"resourceId":"umw-nebulaone-tenant-hosted-ai-platform-capped-usage-20261015","versions":[{"version":"external-e3f9cf4cdc70bc636bfe05e4e49dd4e9cb37e3491539f2c5de1b99277f75015f","resource":{"id":"umw-nebulaone-tenant-hosted-ai-platform-capped-usage-20261015","title":"University of Mary Washington opens NebulaOne, a tenant-hosted multi-model AI platform with capped usage and shared agents, after nearly a year of pilots","organization":"University of Mary Washington (University Communications; IT)","sector":"Public comprehensive university: managed AI platform for staff, faculty and students","geography":"Fredericksburg, Virginia, United States","publishedAt":"15 October 2026 (EagleEye faculty and staff news)","publicationDate":"2026-10-15","eventDate":"2026-10-15","sourceName":"EagleEye, University of Mary Washington","sourceLabel":"Institutional announcement","sourceUrl":"https://eagleeye.umw.edu/2026/10/15/secure-generative-ai-platform-nebulaone-is-now-available/","evidenceClass":"vendor-claim","outcomeClass":"emerging","topics":["infrastructure","knowledge-work","data-security","governance-procurement"],"finding":"After nearly a year of pilots and testing, UMW opened NebulaOne to all students, faculty and staff. The university says prompts stay inside its Microsoft tenant and are not used to train models, and cites privacy, equity of access and AI literacy as reasons. Highly sensitive data is barred; protected or internal data, including FERPA-protected records, may be entered with care. Interactions are not actively monitored, but automated moderation flags dangerous or illegal activity for human review. To manage costs, models are limited and usage is capped monthly, with more by request. Users can build personal and shared agents. A governance committee oversees the platform, which Cloudforce builds on Azure.","sledRelevance":"A smaller public university chose a tenant-hosted, multi-model platform over a single-vendor seat deal, with explicit cost controls through usage caps. That is a different procurement and architecture model from the seat-license announcements in recent editions.","evidence":"University announcement describing scope, rules and governance. No pilot findings, costs, usage numbers or evaluation are reported.","architectureImplications":"The platform runs in the university's own Microsoft and Azure tenant, giving IT control over model selection, quotas and data residency, and responsibility for cost and operations.","governanceImplications":"Dedicated platform governance committee under an AI acceptable use policy; course use stays with instructors.","securityPrivacyImplications":"FERPA records permitted with care; automated moderation with human review; no active monitoring of interactions.","caveats":"Pilot results not shared. Vendor adoption claims in the about section are the university's restatement. Published 15 October, before the last completed run, but not found or archived then; included as a distinct architecture model.","streamIds":["campus-operations"],"roles":{"sales":"Sales framing: UMW shows a tenant-hosted, multi-model option with usage caps, which suits institutions wary of per-seat vendor deals. Bounded help is cost modeling and quota design for a similar platform, or setting up a governance committee. Do not cite pilot results or savings; none were published. The pilot ran nearly a year, so ask for its findings before proposing anything; the announcement does not share them.","engineering":"Engineering framing: the platform sits in the university's Azure tenant, so engineering owns model endpoints, quotas, moderation and shared-agent permissions. Validate that shared agents cannot reach data beyond the creator's rights, that moderation flags route to a named reviewer, and that monthly caps are enforced and reported. Confirm how quota increases are approved and logged, since the help desk handles them by request, and how new models are vetted before the governance committee adds them.","delivery":"Delivery framing: operating this platform means running quota requests, model additions and a governance committee as ongoing services. IT owns operations; the governance committee owns scope. Proposed acceptance: monthly usage and cost reported against caps, and moderation flags reviewed within a set time. The help desk will need a clear rule for approving extra usage or newer models, and the committee needs a regular review of what users ask for."},"roleDetails":{"sales":{"customerProblem":"A smaller university wants broad AI access without per-seat licenses or sending data to consumer services.","stakeholders":["CIO","CFO","AI working group","Faculty senate","Information security"],"discoveryQuestions":["How is Azure consumption budgeted and reported?","Who reviews moderation flags?","What did the pilot find?"],"valueHypothesis":"Interpretation: usage caps may make costs predictable; this source gives no cost data.","potentialEngagement":"Cost and quota model plus governance-committee charter for a tenant-hosted AI platform.","unsupportedClaims":["Cost savings versus seat licenses","Pilot outcomes","Security certification"]},"engineering":{"fit":"Tenant-hosted multi-model AI platform.","architectureIntegration":"Runs in the university's Microsoft and Azure tenant with SSO, model selection, quotas and personal and shared agents.","prerequisites":["Azure subscription and cost controls","Data classification policy","Moderation review process"],"constraints":["Highly sensitive data barred","Limited model selection","Monthly usage caps"],"security":"Data stays in tenant per the university; automated moderation; no active monitoring.","proofOfValue":["Confirm shared-agent permission boundaries","Report usage against caps for one month"]},"delivery":{"work":"Operate quotas, model requests, moderation review and governance for a campus AI platform.","dependencies":["Platform vendor updates","Azure costs","Governance committee"],"ownership":"IT operates; the NebulaOne Governance Committee sets scope.","skillsAdoption":"Basics guide; feedback through the help desk.","governanceCheckpoints":["Governance committee approval for new models","Acceptable use policy compliance"],"acceptanceCriteria":["Proposed: moderation flags reviewed within two business days","Proposed: monthly cost within budget"],"risks":["Cost overrun","Shared agents expose data","Low adoption due to caps"]}},"retrievedAt":"2026-10-19T21:56:40-05:00","enrichedAt":"2026-10-19T22:05:34-05:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Equity rationale of no individual cost; monthly caps may limit heavy users unless they request more.","procurementImplications":"Platform plus cloud consumption model; cost managed through model limits and caps rather than per-seat pricing.","operatingModelImplications":"IT help desk handles quota increases and model access requests; the governance committee manages platform scope.","sourceVerification":{"openedUrl":"https://eagleeye.umw.edu/2026/10/15/secure-generative-ai-platform-nebulaone-is-now-available/","referenceExcerpt":"To manage costs, a limited selection of models is available to all users, and usage of the platform is capped each month.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}