Follow this tutorial that will walk you through setting up a communication system between AI agents using Aiven for Valkey and Amazon Bedrock’s Claude model. The goal is to enable agents to operate independently while staying in sync through a pub/sub service.
This setup is particularly useful in scenarios like automating customer support, managing IoT devices, or any situation where systems need to communicate in real-time.
In this tutorial we'll take an example of two inhabitants of an imaginary planet that meet and start talking.
The terraform files can be found in ./terraform folder of the current project.
./terraform/terraform.tfvars-example provides an example of tfvars file.
- Rename (or create a new file) to
./terraform/terraform.tfvars. - Populate aiven_api_token with the token that you copied from the console.
- Populate project_name with the name of the Aiven project.
- Navigate to
terraformfolder - Set
export PROVIDER_AIVEN_ENABLE_BETA=truein the terminal (Terraform Valkey resource is still in beta stage). - Run
terraform init. - Run
terraform plan. - Run
terraform apply.
Terraform will initiate creation of resources:
- Aiven for Valkey
Once deployment is done, Terraform will also create .env file with necessary credentials to access the services.