Deployment

This section covers how to deploy your Amazon Bedrock Agents to AWS.

Prerequisites

Before deploying your agent, ensure you have:

  • AWS CLI configured with appropriate credentials

  • AWS SAM CLI installed

  • An S3 bucket for deployment artifacts

  • Appropriate permissions to create and manage AWS resources

Deployment Process

The SDK provides a simple way to deploy your agent to AWS using the deploy method:

from bedrock_agents_sdk import Client, Agent

def get_time() -> dict:
    """Get the current time"""
    import datetime
    now = datetime.datetime.now()
    return {"time": now.strftime("%H:%M:%S")}

agent = Agent(
    name="TimeAgent",
    description="An agent that can tell you the current time",
    foundation_model="anthropic.claude-3-sonnet-20240229-v1:0",
    instructions="You are a helpful assistant that can tell the time."
)

agent.add_function(get_time)

client = Client()
client.add_agent(agent)

# Deploy the agent to AWS
client.deploy(
    stack_name="time-agent",
    s3_bucket="my-deployment-bucket",
    s3_prefix="time-agent"
)

Deployment Options

The deploy method accepts several options to customize the deployment:

  • stack_name: The name of the CloudFormation stack to create

  • s3_bucket: The S3 bucket to store deployment artifacts

  • s3_prefix: The prefix for deployment artifacts in the S3 bucket

  • region_name: The AWS region to deploy to

  • profile_name: The AWS profile to use for deployment

  • capabilities: The CloudFormation capabilities to use

  • parameter_overrides: CloudFormation parameter overrides

  • no_confirm_changeset: Whether to skip confirmation of the CloudFormation changeset

  • fail_on_empty_changeset: Whether to fail if the CloudFormation changeset is empty

SAM Template Generation

The SDK generates an AWS SAM template for deployment. You can customize the template generation process using the SAMTemplateGenerator class:

from bedrock_agents_sdk import Client, Agent
from bedrock_agents_sdk.deployment import SAMTemplateGenerator

agent = Agent(
    name="CustomAgent",
    description="An agent with custom deployment",
    foundation_model="anthropic.claude-3-sonnet-20240229-v1:0",
    instructions="You are a helpful assistant."
)

# Create a custom SAM template generator
template_generator = SAMTemplateGenerator(
    memory_size=512,
    timeout=30,
    runtime="python3.9",
    environment_variables={
        "DEBUG": "true",
        "LOG_LEVEL": "INFO"
    }
)

# Deploy with the custom template generator
client = Client()
client.add_agent(agent)
client.deploy(
    stack_name="custom-agent",
    s3_bucket="my-deployment-bucket",
    s3_prefix="custom-agent",
    template_generator=template_generator
)

Updating Deployed Agents

To update a deployed agent, simply make changes to your agent and redeploy:

# Update the agent
agent.instructions = "You are a helpful assistant that can tell the time and date."

# Add a new function
def get_date() -> dict:
    """Get the current date"""
    import datetime
    now = datetime.datetime.now()
    return {"date": now.strftime("%Y-%m-%d")}

agent.add_function(get_date)

# Redeploy the agent
client.deploy(
    stack_name="time-agent",
    s3_bucket="my-deployment-bucket",
    s3_prefix="time-agent"
)