# Use cases

Concrete scenarios that show how Agent Stream’s capabilities combine to address real data protection, safety, cost, reliability, and governance challenges across your AI estate.

| Use case | What it covers |
| [Keep PII safe in every LLM call](/products/affinidi-trust-fabric/agent-stream/use-cases/keep-pii-safe-in-every-llm-call.md) | Catch personal and company data in prompts and responses, redact, pseudonymise, or block it, and ground answers on Microsoft 365 documents without them leaving that boundary. |
| [Shut down the OWASP top risks](/products/affinidi-trust-fabric/agent-stream/use-cases/shut-down-the-owasp-top-risks.md) | Apply one auditable, fail-closed guardrail pipeline across every model and team to block prompt injection, jailbreaks, and toxic or off-topic content. |
| [Stop teams eating the budget](/products/affinidi-trust-fabric/agent-stream/use-cases/stop-teams-eating-the-budget.md) | Cap and attribute AI spend per member, team, and surface, with per-minute rate limits and cost-spike alerts. |
| [Ride out LLM provider outages](/products/affinidi-trust-fabric/agent-stream/use-cases/ride-out-llm-provider-outages.md) | Keep a provider outage from breaking a feature with cross-provider failover and response caching, and switch providers with no redeployment. |
| [Governed models in every editor](/products/affinidi-trust-fabric/agent-stream/use-cases/governed-models-in-every-editor.md) | Roll AI coding assistants out to the whole organisation through corporate SSO, with entitled model menus and per-developer attribution. |
| [Prove what every model did](/products/affinidi-trust-fabric/agent-stream/use-cases/prove-what-every-model-did.md) | Watch live usage, cost, and error rates, and produce a signed, hash-chained Verifiable Credential for every recorded interaction. |

## Pages in this section
- [Keep PII safe in every LLM call](/products/affinidi-trust-fabric/agent-stream/use-cases/keep-pii-safe-in-every-llm-call.md): Stop personal and company data from reaching a model provider, or leaking back out in an answer, from any AI feature you ship.
- [Shut down the OWASP top risks](/products/affinidi-trust-fabric/agent-stream/use-cases/shut-down-the-owasp-top-risks.md): Stop a single missed jailbreak or leaked secret from compromising every AI feature you ship.
- [Stop teams eating the budget](/products/affinidi-trust-fabric/agent-stream/use-cases/stop-teams-eating-the-budget.md): Stop one team's runaway usage from becoming an end-of-month bill nobody can explain.
- [Ride out LLM provider outages](/products/affinidi-trust-fabric/agent-stream/use-cases/ride-out-llm-provider-outages.md): Keep a single provider outage or rate limit from taking your AI features down.
- [Governed models in every editor](/products/affinidi-trust-fabric/agent-stream/use-cases/governed-models-in-every-editor.md): Give every developer AI in their editor without scattering provider keys or losing sight of cost.
- [Prove what every model did](/products/affinidi-trust-fabric/agent-stream/use-cases/prove-what-every-model-did.md): Answer an auditor's "what did the AI actually do" with proof, not a log line they have to trust.

