Ask CloudVision and the AI gateway
Grounded answers from your own tenant data, delivered through a governed, metered AI gateway with any model behind it.
Online version, always current: https://cloudvision360.com/docs/cloudvision360/ask-cloudvision/
Ask CloudVision answers questions about your estate from your own data, and every AI request runs through a gateway you govern.
Open it from Ask CloudVision in the top bar, on every page. Its full page stays reachable from jump-to (Ctrl+K, type “ask”).
Grounded, cited answers
Ask things like “why did the bill go up?” or “which identities can reach production?”. Answers are computed from your tenant’s data first and always cite the findings, resources and policies behind them. Ask CloudVision never invents resources, numbers or causes — if the data cannot support an answer, it says so.
One gateway, any model
Anthropic, OpenAI-compatible APIs, Azure OpenAI, Amazon Bedrock, Google Vertex AI and self-hosted models all sit behind one gateway with routing and fallback. You choose which models are allowed.
Your data stays where policy says
In the hosted service, customer data stays on self-hosted models and never goes to an external AI provider. In a private deployment, data classification and policy decide which providers may see a request, and restricted data never leaves for an external model.
Metered and budgeted
Every token and dollar is metered by tenant, user, feature and model, with quotas, rate limits and monthly budgets. Prompts are registered and versioned, and each answer records the prompt version and a hash of what was sent, written to the audit trail.
No action without policy
AI can only ever recommend actions the policy engine allows. Anything above low risk needs a person, and AI can never approve a change itself.
Related
For external AI provider spend and keys, see AgentVision360. For governing autonomous agents, see AI agent governance.