> ## Documentation Index > Fetch the complete documentation index at: https://docs.manthan.systems/llms.txt > Use this file to discover all available pages before exploring further. # Use cases \# Use Cases Parmana is used anywhere AI systems trigger real-world actions that require authorization, control, and auditability. \--- \## 1. AI Payment Authorization \### Scenario An AI agent proposes a payment: \- vendor payout \- customer refund \- subscription charge adjustment \### Problem Without verification: \- payments may be triggered incorrectly \- fraud risk increases \- no deterministic approval trail exists \### With Parmana \- AI generates payment request \- Governance evaluates authorization policy \- Execution only happens if approved \- Every transaction is cryptographically signed \--- \## 2. AI Infrastructure Changes \### Scenario AI systems manage cloud infrastructure: \- scaling servers \- deploying services \- modifying production configs \### Problem \- accidental deployments \- lack of approval traceability \- unsafe automation loops \### With Parmana \- every change requires authorization decision \- policy defines safe deployment boundaries \- execution is enforced deterministically \--- \## 3. Healthcare Workflow Authorization \### Scenario AI assists in: \- patient escalation \- diagnostic recommendations \- treatment approvals \### Problem \- compliance requirements \- high-risk decisions \- audit requirements \### With Parmana \- every action requires policy-based authorization \- decisions are fully traceable \- audit trails are cryptographically verifiable \--- \## 4. AI Agent Tool Execution \### Scenario AI agents call tools: \- APIs \- external services \- internal systems \### Problem \- agents can trigger unintended actions \- no centralized authorization layer \- unpredictable execution behavior \### With Parmana \- every tool call is validated \- execution requires Authorization Decision \- unsafe actions are blocked deterministically \--- \## 5. Enterprise Automation Systems \### Scenario Enterprises automate: \- finance workflows \- HR actions \- internal approvals \### Problem \- fragmented approval systems \- inconsistent enforcement \- audit complexity \### With Parmana \- centralized deterministic authorization \- unified policy enforcement \- full execution traceability \--- \## 6. AI Agent Platforms \### Scenario Companies building AI agents that: \- take autonomous actions \- interact with APIs \- perform multi-step workflows \### Problem \- lack of safety boundary \- no execution governance \- inability to prove correctness \### With Parmana \- agents generate signals only \- governance decides execution rights \- runtime enforces strict boundaries \--- \## Core insight Across all use cases: > AI proposes actions — Parmana decides whether they are allowed to execute. \--- \## Summary Parmana is required wherever: \- AI systems trigger real-world effects \- execution must be controlled \- compliance and auditability matter \- deterministic authorization is required