AWS

Connecting AWS with Lakehousecat enables Chat and Speech-to-Text (STT) capabilities via Amazon Bedrock and Amazon Transcribe. This page describes how to configure AWS as a Provider Model.

Prerequisites

• Access to Lakehousecat as an Administrator
• AWS account credentials: Access Key ID and Secret Access Key
• The AWS region of your resources

Configuration

Navigate to Workspace → Models → Provider Models and click the + icon.

Step 1: Select Provider Type

From the Provider Type dropdown, select AWS.

Step 2: Select Model Type

Choose the appropriate model type: Chat or STT.

Step 3: Enter Configuration Details

All AWS configurations require the following base fields:

Field	Description
Configuration Name	A unique name for this configuration (e.g. aws-claude-chat).
AWS Access Key ID	Your AWS Access Key ID (e.g. AKIA...).
AWS Secret Access Key	Your AWS Secret Access Key.
AWS Region Name	The AWS region (e.g. us-east-1, eu-west-1).

Chat (Amazon Bedrock)

Field	Description
Inference Profile ARN	The Amazon Bedrock Inference Profile ARN. Retrieve it using the AWS CLI command aws bedrock list-inference-profiles. Example: arn:aws:bedrock:eu-west-1:1234567890:inference-profile/eu.anthropic.claude-3-haiku-20240307-v1:0.

STT (Amazon Transcribe)

Field	Description
STT Model ID / ARN (optional)	Bedrock ARN for Bedrock-based STT. Leave blank for standard Amazon Transcribe.

Step 4: Save

Click Save. After saving, you can optionally set this configuration as the default UI model or default Backend model.

Best Practices

• Use a consistent naming convention that identifies the AWS service and model type.
• Ensure your IAM user or role has the required permissions for Bedrock and Transcribe.

Troubleshooting

• Credential Errors: Verify the AWS Access Key ID and Secret Access Key.
• Region Mismatch: Confirm the region matches where your Bedrock resources are deployed.
• Inference Profile: Use aws bedrock list-inference-profiles to retrieve valid profile ARNs.