RoleWhat bothers them todayWhat Miso gives them
CIO / CTOAI is proliferating across departments with no coherent architecture.One institutional way to reach and govern models, in place of every application building its own stack.
CISO / SecurityEach AI vendor or application adds another data path, credential set, provider relationship, and control surface.One enforcement point where access is restricted, credentials are held centrally, traffic is governed, and activity is logged. Fewer things to secure.
Chief Privacy OfficerDecides again and again whether a particular application may send particular data to a particular model.Privacy decisions that are enforced: who may send what data, for what purpose, to which approved destination.
Legal / ProcurementBAAs, provider terms, and vendor agreements sit next to the software, disconnected from actual AI activity.Agreements linked to use: a route exists because a contract covers that provider, model, purpose, and data. If the agreement does not cover it, the route is not available.
AI / Innovation leaderPromising projects stall in review; teams negotiate the same issues every time.A paved road to production. Once the institution's boundaries are set, new projects inherit them. Approval becomes a capability rather than a project.
Developers / Data scienceFind out which provider is allowed, get credentials, handle PHI correctly, wait for approval, then integrate against provider-specific APIs.One approved endpoint. Build against Miso; the institution's policy decides which models and routes are available.
Vendor management / Third-party riskEvery AI vendor brings its own model stack and opaque downstream dependencies.Vendors route through Miso and inherit the health system's AI controls, in place of selecting models and providers on their own.
Compliance / Internal auditInventory and evidence are assembled after the fact from surveys, spreadsheets, tickets, and interviews.A contemporaneous record of what was approved and what actually ran.
CFO / OperationsDuplicated vendor and model spend, and expensive internal review.Less duplicated infrastructure and review, spend visible by team, route, and vendor, and value from AI investments realized sooner.
Clinical / Operational leaderWants a useful tool deployed and has no interest in model-governance machinery.Less friction between a good use case and production.