Hrida.ai Teams
Hrida.ai Teams
Deploy once, give your entire team access.
Hrida.ai is built to be shared. A single instance can serve your whole organization. Users just open a browser and start chatting. No per-seat installs, no client-side dependencies, no fragmented data across machines.
End users don't need to install anything, manage Docker, or touch a terminal. They open a browser, navigate to your instance URL, and log in.
A shared instance means shared knowledge — channels, shared chats, global prompts, and knowledge bases all available to everyone.
When running local models, a single powerful server (or cluster) serves your entire team instead of requiring capable hardware on every desk.
Manage everything from one place: control which models users can access, configure Role-Based Access Control (RBAC), review audit logs, and restrict features as needed.
A shared instance means shared knowledge.
| Channels | Persistent spaces where your team and AI models work together in real time |
| Shared Chats | Send an exact conversation snapshot to a colleague |
| Global Prompts & Knowledge | Build specialized agents and make them instantly available to everyone |
To share your instance, you need to make it accessible over a network. There are three common approaches, from simple local access to production-grade public domains.
1 Zero-Config LAN (Local Network)
If your team is on the same local network or VPN, they can reach Hrida.ai using your machine's local IP address and port.
2 Secure Private Networks (Recommended)
For remote teams that don't need public internet exposure. Overlay networks and secure tunnels provide encrypted access without opening firewall ports.
Private mesh network with MagicDNS
→Edge exposure with Zero Trust access
→Quick temporary sharing for development or testing
→3 Public HTTPS (Reverse Proxy)
For production deployments, place Hrida.ai behind a reverse proxy for SSL/TLS termination on your own domain (e.g., ai.yourcompany.com).
Industry standard web server and reverse proxy
→Automatic HTTPS with minimal configuration
→High-performance load balancing and proxying
→Once your instance is network-accessible, you need to manage how users create accounts and log in.
The first user to register becomes the Administrator. All subsequent sign-ups are placed in a Pending state and cannot access models or use the platform until an admin approves their account from the Admin Panel.
For organizations where manual approval doesn't scale, Hrida.ai integrates with your existing identity provider.
| OAuth / OIDC | Authenticate via Google, Microsoft, Okta, or Keycloak |
| Group mapping | Map IdP groups directly to Hrida.ai groups |
| SCIM 2.0 | Automated user and group provisioning and deprovisioning |