Getting started with Red Hat Ansible Lightspeed with IBM watsonx Code Assistant

Note from the editor: This blog post was originally published in November 2023 and has been updated to include the latest feature enhancements as well as the addition of a 90-day trial option for Red Hat Ansible Automation Platform subscribers.  

Red Hat Ansible Lightspeed with IBM watsonx Code Assistant is a generative AI (gen AI) service within Red Hat Ansible Automation Platform that is engineered to help automation teams create, adopt and maintain Ansible content more efficiently.

In this post, we’ll walk through the capabilities of Ansible Lightspeed with IBM watsonx Code Assistant to help you turbocharge Ansible content creation experience and start automating faster. We'll also review the options available to help you get hands-on with the service. 

Bringing AI to Ansible content creation through collaboration between Red Hat and IBM

Ansible Lightspeed with IBM watsonx Code Assistant is the culmination of Red Hat and IBM teams combining forces to create a cohesive AI experience for Ansible content creation. Tapping into automation-specific IBM watsonx foundation models, it turns text prompts into Ansible content for the creation of Ansible content. The generated content adheres to accepted Ansible best practices and, when combined with the Ansible code bot, helps teams build more confidence in their automation code base.

The service consists of three components:

1. The developer interface: Built natively into the VS Code extension via the Ansible extension, this allows Ansible content creators to use natural language prompts in the Ansible Playbooks or task files to generate Ansible Lightspeed single and multi-task suggestions.
2. The integrated service: This acts as the glue or broker between the developer interface and the watsonx.ai service. It brings the power of AI to Ansible Automation Platform and enhances the responses from the AI with its post-processing capabilities.
3. Gen AI: IBM watsonx Code Assistant provides access to Ansible-specific watsonx.ai foundation model that generates Ansible content recommendations. This is the “AI guts'' of the solution.

Ansible Lightspeed with IBM watsonx Code Assistant can be deployed as a software-as-a-service (SaaS) deployment or on-premise in your datacenter. 

How to access Ansible Lightspeed

Ansible Lightspeed is included with a subscription to Ansible Automation Platform, however, there is an additional cost to access the IBM watsonx Code Assistant model. 

SaaS-based Service

Ansible Automation Platform customers: start a 90-day trial

Existing Ansible Automation Platform customers can start experimenting with the service through a free 90-day trial to Ansible Lightspeed. Learn how. 

Once your trial expires, explore pay-as-you-go options through an IBM Cloud account. See IBM watsonx Code Assistant pricing plans. 

Not currently an Ansible Automation Platform subscriber? 

Sign up for a trial subscription, then follow these instructions to activate your Ansible Lightspeed trial.  

On-premise Service

For on-premise deployments, you will need to install the Ansible Lightspeed software via the Red Hat OpenShift operator, as well as the IBM watsonx Code Assistant via IBM Cloud Pak for Data. Please contact IBM for pricing and additional information. 

Once you're connected,  it's time to experience its AI-enhanced content creation experience. The following features are available through both the SaaS and on-premise service. 

Create

Create an Ansible Playbook

Ansible Lightspeed with watsonx Code Assistant supports full playbook generation and content explanation. You will need to install the latest version of the Ansible VS Code extension (>=24.8.0) to access these features.

To get started, click on the Ansible icon in the left-hand navigation and then the “Get Started” button under the Ansible content creator. Select “Playbook with Ansible Lightspeed” to enter the guided chat interface for full playbook generation.

In the interface, enter your playbook prompt in natural language and click “Analyze”. The service will provide an editable outline of tasks (pseudo code). This chat-like interface allows real-time interaction with the gen AI service where you can refine your prompt. Both the playbook goal and outline are sent to Ansible Lightspeed with watsonx Code Assistant, which suggests a complete playbook. This intuitive and guided experience helps novice to intermediate users generate full Ansible Playbooks using gen AI.

Content explanation

The Ansible VS Code extension also provides you capability to explain existing and newly generated playbooks so that you can easily understand the details of any playbook. To explain any playbook, open it in VS Code and click on “Explain the current playbook” under the Ansible Lightspeed section. You can also right-click on a playbook and click on “Explain the playbook with Ansible Lightspeed”.

Full playbook generation with explanation

Along with full playbook generation and playbook explanation, Ansible Lightspeed also provides more granular control over the creation process through features like single and multi-task generation.

Single and multiple task generation

When you need to add specific inline tasks to your playbooks, Ansible Lightspeed excels with its single and multi-task generation capabilities.

Single Task Generation

• Start a new task in your playbook and provide a name using the name argument.
• Position your cursor at the end of the task name and hit “Enter” to get a suggestion from Ansible Lightspeed.

Multi-task Generation

• Create a comment line at the task block level in your playbook.
• Chain multiple natural language prompts together using ampersands (&).
• At the end of the comment line, hit “Enter” to get a multi-task suggestion from Ansible Lightspeed.

These features streamline the process of creating detailed, multi-step workflows in your Ansible Playbooks, enhancing efficiency and consistency.

Generating an Ansible task
 

Generating multiple Ansible tasks

Adopt

Content source matching

Another critical feature of Ansible Lightspeed with watsonx Code Assistant is transparency and openness using content source matching. We transparently share the potential source, author and content license of the training data used for the recommendation. Building trust in the community and supporting the relationships between authors and contributors is part of Red Hat’s DNA.

Ansible Lightspeed content source matching

Post-processing

Ansible Lightspeed is purpose-built for automation, providing a streamlined and accurate gen AI experience that is tailored by—and for—IT automation teams. Ansible Lightspeed's post-processing capabilities adhere to accepted Ansible best practices so automation teams can have confidence in the code they are accepting.

Examples of post-processing include using FQCN, doing data anonymization and other variable substitutions. The example provided in the “How to generate a suggestion” section above shows the suggestions generated with FQCNs as one of the post-processing capabilities of the service.

Maintain

Model customization (fine-tuning)

Model customization allows Ansible Lightspeed users to use their existing Ansible content to train the model. This feature can help to improve the quality and accuracy of your Ansible content with code recommendations that are more tailored to your organizations' specific needs and automation patterns.

At a high level, there are two steps to this process. The first step is to create a training dataset in a format that can be consumed by IBM watsonx Code Assistant to create a custom model. The second step is to feed the training data to the gen AI model in the IBM Cloud console through the Watson Studio on IBM watsonx. Sticking to the collaborative approach between Red Hat and IBM, Red Hat has created a command-line tool to cover the first step of the process called ansible-content-parser.

Ansible code bot

Customers often ask, “we have the code, but what's next”? How do we move from siloed automation? How do we create a community of practice? Our answer starts with culture, and treating infrastructure workflows as software assets catalyzes the collaborative approach needed for enterprise-level automation. However, this could involve learning new skills like version control, automated testing and code review.

Our goal with Ansible code bot is to extend Ansible content quality improvement beyond creation and into your entire workflow. The Ansible code bot scans existing Ansible Content Collections, roles and playbooks hosted in GitHub repositories, and proactively creates pull requests whenever best practices or quality improvement recommendations are available. The bot automatically submits pull requests to the repository, which proactively alerts the repository owner to a recommended change to their content. You can configure Ansible code bot to scan your existing Git repositories (both public and private).

Prerequisites

• A Red Hat customer portal account that is attached to your organization.
• Your organization should have a valid Ansible Lightspeed subscription enabled.

Procedure

1. Log in to your GitHub by using an account associated with your organization.
2. Install the GitHub app for the organization that you are a member of.
3. Go to the Ansible code bot GitHub app:
   1. Ansible code bot - GitHub Apps
4. Select the repositories on which you want to install the Ansible code bot.
5. Click Install & Authorize.
6. When prompted, log in to your Red Hat SSO account.

Ansible code bot in action

After the Ansible code bot is installed for the Git repositories you selected, you can configure a schedule to scan your Git repositories at regular intervals. You can also manually scan your Git repositories by adding a topic called “ansible-code-bot-scan” to your repository, if you have not set up a scanning schedule for your Ansible code bot or if you do not want to wait for the next scheduled scan.

Additional resources

• Website: Red Hat Ansible Lightspeed with IBM watsonx Code Assistant
• Documentation: Ansible Lightspeed with IBM watsonx Code Assistant
• Datasheet: Ansible Lightspeed with IBM watsonx Code Assistant

Thanks for reading and happy automating.