Generative AI demands infrastructure that’s not only powerful but repeatable, auditable, and scalable. From chat bots and content generation to intelligent automation agents, organizations are deploying AI at scale. But with this innovation comes complexity. In short, deploying generative AI isn’t just about models, it’s about managing the infrastructure and operations behind them reliably. The Red Hat Certified Collection, amazon.ai, addresses this problem by bringing infrastructure-as-code principles to AI and operational monitoring.
The problem: Manual AI management doesn’t scale
Even with powerful services like Amazon Bedrock and DevOps Guru, organizations face hurdles:
- Agent lifecycle complexity: Bedrock Agents can orchestrate multiple models and APIs, but creating, updating, and validating these agents manually is tedious and prone to error.
- Operational blind spots: Monitoring resources and enabling compliance using DevOps Guru often involves manual configuration and report generation, which is not only slow but inconsistent.
- Non-reproducible deployments: Without automation, recreating an agent or a monitoring setup in another environment may yield subtle differences, leading to unexpected failures or audit issues.
- Time and resource drain: Engineers spend hours repeating routine configuration steps instead of building and improving AI-powered applications.
Part of the problem is that AI and the infrastructure it requires is complex. There are many distinct parts, including foundation models, runtime endpoints, AI agents, and orchestration logic, and they must all be deployed and maintained. Manual setup in dev, staging, and production can easily become inconsistent. And updating models or agent logic often involves repetitive clicks in AWS consoles, which slows delivery and increases risk. For regulated industries, it's also essential to track who changed what, and when.
Introducing the amazon.ai Collection
The Red Hat Ansible Certified Collection, amazon.ai, brings infrastructure-as-code principles to AI and operational monitoring. Think of it as the automation bridge between AI innovation and enterprise-grade operational discipline. Say goodbye to manual clicks in the console and hello to idempotent, automated AI deployments:
- Idempotent automation: Run the same playbook multiple times and achieve the same configuration without errors.
- Audit-ready: Capture agent configurations, model versions, and operational settings programmatically, producing structured audit logs.
- Scalable: Apply consistent AI and monitoring setups across hundreds of environments with minimal effort.
- Integrated: Seamlessly ties Amazon Bedrock and DevOps Guru into existing Ansible pipelines and AWS infrastructure.
The Red Hat AnsibleCertified Collection for amazon.ai is a new set of Ansible modules purpose-built for Amazon Bedrock and DevOps Guru, allowing engineering, MLOps, and platform teams to define, deploy, and manage generative AI infrastructure as code.
Core capabilities of Red Hat AnsibleCertified Collection for amazon.ai
The Red Hat AnsibleCertified Collection for amazon.ai is designed to bring idempotent, auditable, repeatable automation to your AI stack, just like you already have for EC2, S3, or ECS. With this collection, you can deploy and update Bedrock Agents programmatically, configure action groups and aliases without touching the console, invoke models and agents directly from playbooks, and automate DevOps Guru monitoring, anomaly detection, and audit reporting.
Automating generative AI with Amazon Bedrock
Amazon Bedrock is a fully managed service that offers a choice of high-performing foundation models (FMs) along with a broad set of capabilities for building and scaling generative AI applications. With Bedrock’s Agents, developers can orchestrate tasks using multiple models and APIs, and with the Bedrock Runtime, they can invoke models directly for inference.
Amazon Bedrock lets you deploy foundation models quickly, but scaling or maintaining agents across environments can get messy. The Red Hat AnsibleCertified collection for amazon.ai brings the power of infrastructure-as-code to AI workflows so you can deploy, update, and validate agents automatically, every time.
Bedrock Agent management modules
These modules provide complete control over the lifecycle of a Bedrock Agent, including its core configuration, abilities, and deployment endpoint.
bedrock_agent: Manages the creation, updating, and deletion of the core Bedrock Agent, including its foundation model, IAM role, and orchestration instructions.bedrock_agent_info: Retrieves detailed information about one or all Bedrock Agents.bedrock_agent_action_group: Creates, updates, and deletes Action Groups for a Bedrock Agent, defining the operational capabilities (Lambda function/OpenAPI schema) the agent can use.bedrock_agent_action_group_info: Retrieves detailed configuration for specific or all Action Groups for a given agent version.bedrock_agent_alias: Manages the creation and deletion of aliases, which provide a stable, versioned endpoint for invoking a specific agent version.bedrock_agent_alias_info: Fetches details for specific or all aliases associated with an agent.bedrock_foundation_models_info: Lists and filters available foundation models in Bedrock based on criteria like provider, customization type, or output modality.
Bedrock runtime modules
These modules facilitate direct interaction with the deployed agents and models for testing and operational integration.
bedrock_invoke_agent: Interacts with a deployed Bedrock Agent by sending a text prompt, maintaining session history, and optionally tracing the agent’s reasoning process.bedrock_invoke_model: Run direct inference on a foundation model with fine-grained control over inputs and outputs.
Managing operational performance with DevOps Guru
Amazon DevOps Guru is an ML-powered service that automatically detects operational anomalies and insights across your applications. It uses machine learning models to identify deviations from normal operating patterns and provides actionable recommendations.
Maintaining operational visibility and compliance across AWS resources can be challenging, especially when manual configuration is involved. With Amazon DevOps Guru and the Red Hat AnsibleCertified collection for amazon.ai, you can automate resource monitoring configuration, high-severity alert notifications, diagnostics and post-mortem reporting from Red Hat Lightspeed, and audit report generation so your operations are consistent, compliant, and easy to audit.
DevOps Guru modules
These modules help your operational monitoring to be consistently configured, and allow you to pull crucial diagnostic data programmatically.
devopsguru_resource_collection: Manages the set of AWS resources that DevOps Guru monitors, based on AWS CloudFormation stack names or resource tags. It also configures notification channels (SNS).devopsguru_resource_collection_info: Fetches and displays the current configuration of the DevOps Guru resource collection for a specified type (e.g., CloudFormation or Tags).Devopsguru_insight_info: Retrieves detailed information about Proactive and Reactive insights, including related anomalies, affected resources, and the recommended remediation steps.
Why Red Hat AnsibleCertified Collection for amazon.ai matters to you
Automation isn't new, but applying it to generative AI infrastructure is a significant step forward. All too often, AI teams and operations teams work in separate silos. This collection allows both to speak the same language through declarative playbooks. Developers can focus on building new AI experiences instead of babysitting infrastructure.
This supports compliance and governance by allowing you to generate structured reports and to keep a historical record of deployments and changes. This is crucial in regulated sectors like finance, healthcare, or government.
Idempotent and repeatable automation ensures consistency, and minimizes configuration errors. In short, this collection solves the key pain points that prevent AI at scale from being practical in real-world enterprise environments.
Get started with the Red Hat AnsibleCertified Collection for amazon.ai
To start using the collection, make sure your environment is ready with the following requirements:
- ansible-core 2.17 or above
- Python 3.8 or above
- boto3 and botocore 1.35 or above
The amazon.ai Certified collection is available on Red Hat Automation Hub. To consume this Collection from Ansible automation hub, add this to your ansible.cfg file:
[galaxy]
server_list = automation_hub
[galaxy_server.automation_hub]
url=https://cloud.redhat.com/api/automation-hub/
auth_url=https://sso.redhat.com/auth/realms/redhat-external/protocol/openid-connect/token
token=<SecretToken>The token can be obtained from the Ansible automation hub web UI.
Once the above steps are complete, run the command to install the collection:
$ ansible-galaxy collection install amazon.aiYou're ready to automate!
Automation for your AI infrastructure
The launch of the Red Hat Ansible Certified collection for amazon.ai is more than just adding new modules. It closes the loop between AI innovation and enterprise operations. By bringing idempotent and declarative automation to Amazon Bedrock and DevOps Guru, engineering teams can finally manage, validate, and audit their AI infrastructure with the same reliability as traditional cloud services.
Are you new to Ansible automation and want to learn?
- Visit us at the Red Hat booth at AWS re:Invent 2025
- Check out Red Hat Summit 2025
- For further reading and information, visit other blogs related to Red Hat Ansible Automation Platform
- Check out the YouTube playlist for everything about Ansible Collections
- Check out our getting started guide on developers.redhat.com
- Check out the Amazon Web Services Guide
- Try out the hands-on interactive labs
- Read the e-book: Using automation to get the most from your public cloud
关于作者
