> ## Documentation Index
> Fetch the complete documentation index at: https://docs.beyondguard.ai/llms.txt
> Use this file to discover all available pages before exploring further.
# The BeyondGuard Platform: Your AI Control Plane
> BeyondGuard's control plane centralizes model management, policy enforcement, security testing, and telemetry for enterprise AI — all behind role-based access and your own identity provider.
The BeyondGuard platform is an enterprise **AI control plane**: a single place to standardize, secure, and govern how your organization uses AI. It brings dynamic model management across multiple LLM providers, real-time security and compliance controls, RAG-based knowledge layers, role-based access, and detailed operational telemetry together under one orchestration framework.
Everything is built to comply with **OWASP, MITRE, NIST, GDPR/KVKK, and HIPAA** requirements, so AI can be operated securely and audited end to end.
## What's in the control plane
Real-time security posture, anomaly detection, and usage analytics.
The AI Proxy and Context Proxy that inspect every request and response.
Versioned, group-scoped policy sets assigned per endpoint.
A central inventory of every model in use, local or cloud.
Security-scan models before they reach production.
Discover unauthorized third-party AI usage on your network.
Watermarked, fully traceable audit logs and SIEM export.
Multiple governed AI assistants on one platform.
## Roles
Access is role-based, and each role determines the level of access to operational and administrative controls.
| Role | Scope |
| - | - |
| **Super Admin** | The broadest privileges across the whole platform — create organizations, define security policies, manage proxy configurations, change user roles, and view all logs. |
| **Admin** | Administrative authority within their own organization — add and edit users, configure organization settings, manage security controls, and review operational logs. |
| **User** | Runs operational queries within the organization's defined scenarios; can view only their own usage, with no access to administrative functions. |
An organization is defined first (under **Settings → Organization** by a Super Admin), then users are created and assigned roles in **User Management**.
## Identity and multi-tenancy
BeyondGuard integrates with your existing identity infrastructure — **LDAP, Active Directory, and SSO** (SAML / OAuth providers) — as well as endpoint Bearer tokens. Users sign in with their existing corporate accounts, and their **organization is carried over automatically at authentication**: each user is directed only to the policies, projects, models, and logs belonging to their own organization. This is what makes BeyondGuard's **multi-tenant** model safe — different organizations, business units, and partners operate on one platform with fully isolated data.
## Related
The protection model the platform enforces.
Roll the platform out across your AI stack.
Integrate applications with the proxy.
How the platform is built and deployed.
This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.