Red Hat introduces Ansible Lightspeed to power AI-powered IT automation

Red Hat introduces Ansible Lightspeed to power AI-powered IT automation

Join top executives in San Francisco July 11-12 to hear how leaders are integrating and optimizing AI investments for success. Learn more

Open source solutions company Red Hat today introduced Ansible Lightspeed, a generative AI service integrated with IBM Watson code assistant. The company’s latest offering seeks to drive widespread use of Ansible automation within organizations by making task automation easier for beginners and freeing experienced automaters from the daunting task of building low-level tasks.

Red Hat uses natural language processing (NLP) to integrate the service with Watson Code Assistant, which is expected to be available in the near future. Ansible Lightspeed allows users to quickly build automation code by leveraging IBM’s core templates. According to the company, this integration offers a valuable solution for enterprises, as it closes the skills gap and improves efficiency, thus accelerating the time-to-value of automation.

Continuous feedback

By allowing users to enter simple prompts in English, the service makes it easy to translate domain expertise into YAML code to create or edit Ansible Playbooks. Furthermore, users can actively contribute to the formation of the model by providing valuable feedback and ensuring continuous improvements.

“Organizations looking to modernize face a key challenge: an automation skills gap,” Tom Anderson, vice president and general manager of Ansible, told VentureBeat. “Generative AI has the potential to make savvy automation talents more productive and expand the openness of those who can create actionable automation content.”


Transform 2023

Join us in San Francisco July 11-12, where top executives will share how they integrated and optimized AI investments for success and avoided common pitfalls.

subscribe now

As? By making it easier for automation domain experts to translate their expertise into working automation code, she said. “Users can use natural language prompts to get code recommendations for task generation, which are the building blocks of Ansible Playbook,” Anderson said.

The company says this new tool enables domain experts to effectively translate process knowledge and goals into code. It is also aimed at users who have a deep understanding of what needs to be accomplished but lack the YAML expertise to independently create compliant and efficient playbooks.

In addition, the tool leverages Ansible’s extensive repository of subject matter expertise within the Ansible Lightspeed core model. This allows users to explore new automation domains.

Leverage natural language templates to simplify automation

Anderson told VentureBeat that the NLP-optimized base model is the cornerstone of the collaboration between Red Hat and IBM, setting Ansible Lightspeed apart from other tools.

“The base model is trained with data from Ansible Galaxy, a massive open source repository of Ansible content that covers a plethora of use cases and vertical applications of Ansible technology,” said Anderson. “In addition to the data from Ansible Galaxy, the model was (and continues to be) fine-tuned with additional IT automation expertise from Red Hat and IBM.”

He said he believes IT automation is a key driver of operational efficiency and allows teams to focus on innovation. However, managing automated workflows can be complicated and time consuming. Ansible Lightspeed can increase the efficiency of an organization’s automation efforts and improve ROI and time-to-value.

“Writing quality automation code takes time and resources,” Anderson said. “Ansible Lightspeed can help developers and operations teams produce better automation code much faster. Again, Ansible Lightspeed is not meant to be a silver bullet. But it’s a real enhancement to the crafting experience.

He added that users can access the service directly in their code editor for a “real-time productivity boost” to their existing workflows. “How much time you’re saving depends on the complexity of the playbooks you’re developing, but when you reduce a 30- to 60-minute activity to 5 or 10 minutes, multiple times a day, it adds up,” she said.

Leveraging IBM Research LLM

According to the company, the development of the tool involved the use of a Large Language Model (LLM) derived from IBM Research. IBM contributed its LLM expertise, while Red Hat contributed its specific domain knowledge to train the model using publicly available Ansible automation content.

This collaborative effort also included subject matter experience-based post-commendation training, highlighting the combined strengths of Red Hat’s domain experience and IBM’s expertise in LLM, business modeling, and AI.

“This uses Generative AI from IBM’s core model trained on a specific domain (Ansible), to help people build automation faster,” Anderson explained. “Current subject matter experts will be much more efficient compiling many repetitive pieces of code while creating an automation playbook, which is ultimately YAML code. This accelerates their ability to generate it much faster. Ansible Lightspeed makes existing subject matter experts much more efficient by doing much of the work for them.”

Anderson added that the IBM CIO team was actively involved as the first tester of Ansible Lightspeed with IBM Watson Code Assistant, resulting in dramatic productivity gains.

Continuous accessibility tailored to IT environments

Additionally, during the pilot phase, the preview version of Watson Code Assistant proved instrumental in helping IBM CIO teams generate approximately 60% of their code accurately as they adopted the Ansible automation platform.

The tool offers users accessibility through the Ansible VSCode extension, allowing them to directly interact with the AI ​​within their code editor. Users can solicit the AI, evaluate suggestions and make changes or accept/reject them, with the convenience of embedding the generated code in an Ansible Playbook.

In addition, Ansible Lightspeed works within your IT environment, gaining knowledge and providing recommendations on variables and settings tailored to meet specific requirements.

Additionally, the tool boasts pre- and post-processing capabilities, ensuring that all code hints align with recognized best practices in Ansible and automation. This feature allows users to confidently leverage Generative AI, knowing that recommendations adhere to established guidelines and standards.

“All generated code recommendations are supported by ‘content source matching’,” Anderson said. “This means that users can see the specific URL and location where the code was pulled from, a description of the data source, the license under which the code is covered, and the type of Ansible content it is. All users of Ansible Galaxy will be able to choose not to use their code as data to train the base model of Ansible Lightspeed.

The promising future of automation powered by generative AI

Anderson said Red Hat recognizes the potential of foundational models to deliver significant business value.

Data scientists and developers can improve accuracy by tailoring these models for specific use cases such as writing automation code. However, initial training of these models requires considerable infrastructure and resources, including specialized tools and platforms, before even tackling service, optimization, and management.

These are challenges Red Hat OpenShift AI can help address by providing a foundation that is already familiar to enterprise IT organizations that need to manage AI infrastructure and can still meet the needs of data scientists and app developers, Anderson said.

“We see domain-specific AI as a key driver for future adoption – taking a model and shaping it to meet a specific need for an organization is incredibly valuable,” he said. “This helps build unique AI-enabled applications, and with a foundation like OpenShift AI, you can run them on a manageable and scalable platform that further fuels innovation.”

He explained that the company aims to expand the accessibility of artificial intelligence to enterprises through hybrid cloud.

“This could (include) organizations that want to use the base models internally but need a platform they can use, or it could be a business that just wants to reap the benefits of an AI-powered application without dealing with any plumbing ,” He added. “Red Hat’s goal is to support both of these paths through open standards-based approaches and give customers choice: choice of tools, choice of method of implementation, and choice of how they consume the end product.”

VentureBeat’s mission it is to be a digital city square for technical decision makers to gain insights into transformative business technology and transactions. Discover our Briefings.