LLM Management

Introduction

This documentation provides comprehensive guidance on using the LLM Management dashboard, which allows super administrators to add, configure, and manage language models from various providers. The LLM Management dashboard is a critical component for customizing available AI models within the platform.

Table of Contents

• LLM Management
  • Introduction
  • Table of Contents
  • Accessing LLM Management
    • Navigation Path
    • Access Restrictions
  • Dashboard Overview
  • Provider Management
    • Expanding Provider Options
    • Adding Models to Library
    • Configuring Provider Visibility
  • Model Library Management
    • Viewing Added Models
    • Foundation Models
  • Permissions and Access Control
  • Troubleshooting
    • Common Issues
    • Support Resources
  • Next Steps

Accessing LLM Management

Navigation Path

1. Log in to the platform with super admin credentials
2. Locate the left navigation bar
3. Select Global Settings from the navigation options
4. From the displayed options, click on LLM Management

Access Restrictions

Important: LLM Management is accessible only to users with super admin privileges. Regular admin users or standard users will not see this option in their navigation menu.

Dashboard Overview

Once you access the LLM Management section, you will see the main dashboard interface with the following key components:

Component	Description
Provider Section	Lists all available LLM providers (e.g., OpenAI, Anthropic, etc.)
Configuration Controls	Buttons to enable/disable and configure provider settings
Model Library Tabs	Tabs to switch between available models and your added models
Action Buttons	Buttons for adding models and other management functions

Provider Management

Expanding Provider Options

To view available models from a specific provider:

1. Locate the provider in the list (providers are typically organized alphabetically)
2. Click on the dropdown arrow next to the provider name
3. The dropdown will expand to reveal available models and configuration options

Adding Models to Library

To add a model from a provider to your model library:

1. Expand the dropdown for the desired provider
2. Locate and click the Add to Model Library button (appears as a "+" icon with text)
3. A dialog box will appear requesting necessary credentials
4. Enter the required API key for the selected provider
5. Select the appropriate user group that should have access to this model
   • Groups determine which users can access the model once added
6. Click the Add button to complete the process

API Key Format Example:

OpenAI: sk-xxxxxxxxxxxxxxxxxxxxx
Anthropic: sk-ant-xxxxxxxxxxxxxxxxxxxxx

Note: Ensure you're using a valid API key with sufficient permissions and quota for the model you're adding.

Configuring Provider Visibility

As a super admin, you can control which providers are visible to other administrators:

1. Locate the Configuration button next to the provider
2. Click the button to open provider visibility settings
3. Toggle the visibility option to enable/disable the provider
4. Confirm your changes

When a provider is disabled:

• It will not be visible to other admin users
• Models from this provider will not appear in selection options throughout the platform
• Existing implementations using these models will need alternative configurations

Model Library Management

Viewing Added Models

To view all models that have been added to your platform:

1. Click on the My Model Library tab at the top of the LLM Management dashboard
2. This view displays all models that have been successfully added to your platform

The My Model Library view provides:

• A comprehensive list of all added models
• Details about each model including provider and configuration status
• Options to manage existing models

Foundation Models

The My Model Library includes both:

• Custom Added Models: Models you've added with your own API keys
• Foundation Models: Pre-configured models available to all administrators

Foundation models are:

• Pre-integrated into the platform
• Available without requiring additional API keys
• Managed by the platform provider
• Subject to platform licensing restrictions

Permissions and Access Control

LLM Management provides granular control over which user groups can access specific models:

Access Level	Description
Super Admin	Can add, configure, and manage all aspects of LLM providers and models
Group-Based Access	Models can be restricted to specific user groups during the addition process
Visibility Control	Entire providers can be hidden or shown to different administrative users

Troubleshooting

Common Issues

Issue	Possible Solution
API Key Invalid	Verify the API key format and ensure it has not expired
Model Not Appearing in Library	Refresh the page or check if the model was added to the correct group
Provider Configuration Failed	Ensure your account has access to the specific model with the provider
Access Denied	Verify you have super admin privileges required for LLM Management

Support Resources

If you encounter persistent issues with the LLM Management dashboard:

• Check the platform documentation for updated instructions
• Contact your platform provider's support team
• Verify your license includes access to custom LLM configuration

Next Steps

After configuring your LLM Management settings:

• Test the added models in your applications
• Configure model-specific parameters in your implementation
• Monitor usage and performance metrics
• Adjust configurations as needed based on performance and cost

This documentation is intended for super administrators responsible for configuring and managing language models within the platform. For standard user guides or implementation documentation, please refer to the relevant sections of the platform documentation.