AI Service Providers

The AI Service Providers section allows administrators to configure and manage the AI engines used by IoTRoutes.

A Service Provider represents an AI Agent capable of executing one or more AI capabilities inside the platform.

IoTRoutes supports both:

• cloud-based AI providers,
• and locally hosted AI models.

This architecture allows organizations to:

• distribute AI workloads,
• isolate sensitive operations,
• optimize operational costs,
• and select the best model for each business scenario.

Accessing AI Providers

Navigate to:

AI Configuration → Service Providers

From this section, administrators can:

• create new AI providers,
• configure provider endpoints,
• assign supported capabilities,
• enable or disable providers,
• manage multiple AI backends simultaneously.

Supported Provider Types

IoTRoutes can integrate with multiple AI providers, including:

• OpenAI
• Azure OpenAI
• Ollama
• Custom AI endpoints compatible with OpenAI APIs

The platform is designed to be extensible and can support additional providers in future versions.

AI Providers List

The Service Providers page displays all configured AI providers.

Each provider entry contains:

Administrators can quickly identify:

• available providers,
• assigned capabilities,
• active/inactive services,
• and execution routing options.

Creating an AI Service Provider

To create a new provider:

1. Open the Service Providers page.
2. Click Add new.
3. Configure the provider settings.
4. Assign the allowed capabilities.
5. Enable the provider.
6. Save the configuration.

Provider Configuration

Each AI provider contains the following settings.

Service Name

A unique display name used to identify the provider inside IoTRoutes.

Examples:

• OpenAI
• AzureOpenAI
• Ollama-Local
• Internal-AI

Model

The AI model used by the provider.

Examples:

• gpt-4
• gpt-4.1-mini
• llama3
• mistral
• custom-model

The selected model determines:

• reasoning capabilities,
• response quality,
• execution speed,
• and operational costs.

Endpoint

The API endpoint used to communicate with the AI provider.

Examples:

• OpenAI API endpoint
• Azure OpenAI endpoint
• Local Ollama server
• Custom enterprise AI gateway

IoTRoutes communicates with providers using API-based integrations.

API Key

Authentication key used to access the AI service.

This field is optional for some local providers but required for most cloud-based providers.

For security reasons:

• API keys should be protected,
• rotated regularly,
• and restricted using provider-side permissions whenever possible.

Service Capabilities

Each provider can execute only the capabilities explicitly assigned to it.

This is one of the core security and governance concepts of the IoTRoutes AI Layer.

Examples of capabilities:

• Chat
• IoTMessageToPMS
• SummarizeIoTActivities
• PredictWithLLM

This mechanism allows:

• workload separation,
• provider specialization,
• secure execution boundaries,
• local-only processing for sensitive tasks.

Example architecture:

Enabling or Disabling a Provider

Providers can be enabled or disabled at any time.

Disabled providers:

• remain configured,
• but are excluded from AI execution workflows.

This is useful for:

• maintenance operations,
• provider migration,
• testing environments,
• temporary failover scenarios.

Architecture Benefits

The AI Service Provider architecture provides several advantages.

Multi-Provider Strategy

Organizations can use multiple AI providers simultaneously.

Benefits:

• high availability,
• cost optimization,
• model specialization,
• reduced vendor dependency.

Security Isolation

Sensitive operations can run exclusively on local providers.

Example:

• internal operational analysis executed using Ollama hosted inside the company infrastructure.

This reduces exposure of sensitive operational data.

Capability Routing

AI tasks can be routed dynamically depending on:

• provider capabilities,
• model specialization,
• security requirements,
• or infrastructure policies.