Use this setup to connect OpsPilot to your own LLM provider and enable AI-powered capabilities in OpsRamp.

This enables you to use your own enterprise-managed LLMs while still benefiting from OpsPilot’s conversational and analytical capabilities.

You can connect through two access types:

• DIRECT_API - Direct API access (OpenAI only)
• CLOUD_BASED - Cloud provider access (Vertex AI, AWS Bedrock, Azure Foundry)

Default State

• OpsPilot is disabled by default until an administrator configures valid LLM credentials.
• Once configured and saved, OpsPilot is automatically enabled for your account.

Prerequisites

Before configuring OpsPilot with your own LLM, ensure the following:

• OpsRamp platform access with appropriate permissions to manage credentials
• Valid credentials for your chosen LLM provider
• Models enabled or deployed in your cloud account

Google Cloud Platform Requirements

• A valid GCP project with billing enabled
• Vertex AI API enabled in the project
• Enable the gemini-2.5-flash-lite model in Google Vertex AI, as this model is used for OpsPilot guardrails
• Appropriate IAM permissions to create/manage service accounts

Service Account Requirements

Create a service account and ensure it has the following minimum roles:

You must download the JSON key for this service account. This key is pasted into OpsRamp.

AWS Bedrock Requirements

• A valid AWS account with Bedrock access enabled in the target region
• IAM credentials (access key and secret key) with permissions to invoke Bedrock models
• The required model IDs enabled for your account in Bedrock

Azure Foundry Requirements

• An Azure AI/OpenAI resource endpoint
• A valid API key for that resource
• Supported model deployments available in your subscription

OpenAI Requirements

• A valid OpenAI API key
• Access to the required model IDs for the selected tier(s)

Supported LLM Providers

These models are typically used in the following tiers:

• Standard Tier: Commonly mapped to Flash Lite or Flash, suited for cost-effective, general-purpose workloads.
• Advanced Tier: Commonly mapped to Flash and Pro (or another higher-capability configuration) and designed for:
  • Complex reasoning scenarios
  • Use cases requiring large context windows
  • Deep analysis across extensive infrastructure and observability data

To ensure OpsPilot can effectively support OpsRamp use cases, at least one configured model must meet the minimum requirements for advanced reasoning and large-context analysis.

OpsPilot Model Tiers: Standard vs Advanced

OpsPilot supports separate Standard and Advanced model tiers to ensure the right balance between performance, cost efficiency, and analytical depth. Each tier is optimized for different investigation scenarios.

OpsPilot automatically selects which tier to use for each request. Based on the complexity of the plan the agent builds (for example, whether it needs multi-step reasoning, broader context, or deeper correlation across alerts/tickets/metrics), OpsPilot will route the request to either the Standard or Advanced model tier.

How to Enable OpsPilot Using Your Own LLM

To create a new LLM credential:

1. Navigate to Setup → Account → Credentials.

2. Click +ADD.
   The ADD CREDENTIAL page is displayed.

   

3. In the Credential Details form, set Credential Type to LLM.

4. Enter a Name and optional Description.

5. Select your Access Type (CLOUD_BASED or DIRECT_API).

6. Complete the provider-specific fields as described in the following sections.

7. Select the Standard Tier Model ID (required) and optionally the Advanced Tier Model ID.

8. Click Save.

Credential Details

The following table summarizes the common credential fields:

Configure Vertex AI (Cloud-Based)

To configure a Vertex AI credential, select Access Type as CLOUD_BASED and Cloud Provider as VERTEX_AI.

Required Fields

Optional Fields

• Advanced Tier Model ID - Select from dropdown for advanced queries

Configure AWS Bedrock (Cloud-Based)

To configure an AWS Bedrock credential, select Access Type as CLOUD_BASED and Cloud Provider as AWS_BEDROCK.

Required Fields

Optional Fields

• Advanced Tier Model ID - Enter model ID for advanced queries

Note: For AWS Bedrock, the Standard Tier Model ID and Advanced Tier Model ID are text input fields, not dropdowns. You must manually enter the full model identifier.

Configure Azure Foundry (Cloud-Based)

To configure an Azure Foundry credential, select Access Type as CLOUD_BASED and Cloud Provider as AZURE_FOUNDRY.

Required Fields

Optional Fields

Configure OpenAI (Direct API)

To configure an OpenAI credential, select Access Type as DIRECT_API and Provider as OPENAI.

Required Fields

Optional Fields

Model Tier Selection

Each credential requires a Standard Tier Model ID and optionally supports an Advanced Tier Model ID:

• Standard Tier Model ID (Required) - Default model used for all OpsPilot interactions.
• Advanced Tier Model ID (Optional) - Separate model for complex or advanced queries.

Important: For all providers except AWS Bedrock, the model ID fields are dropdowns showing available models. For AWS Bedrock, you must manually enter the full model identifier.

Once saved, OpsPilot will switch from disabled to enabled.

Completing the Setup

After saving your credentials:

• OpsPilot becomes enabled.
• Search, Command Center, and PRC start using this LLM.
• No additional activation, restart, or configuration is required.